| *** baoli has quit IRC | 00:02 | |
| *** achanda has joined #openstack-nova | 00:06 | |
| *** yuntongjin has joined #openstack-nova | 00:17 | |
| *** ssurana has joined #openstack-nova | 00:21 | |
| *** oomichi has joined #openstack-nova | 00:21 | |
| *** promulo__ has joined #openstack-nova | 00:21 | |
| *** promulo has quit IRC | 00:25 | |
| *** rdo has quit IRC | 00:27 | |
| *** rdo has joined #openstack-nova | 00:29 | |
| *** meghal has joined #openstack-nova | 00:31 | |
| *** meghal1 has joined #openstack-nova | 00:33 | |
| *** haomaiwang has quit IRC | 00:35 | |
| *** meghal has quit IRC | 00:36 | |
| *** ssurana has quit IRC | 00:39 | |
| *** saikrishna has joined #openstack-nova | 00:40 | |
| *** takedakn has quit IRC | 00:42 | |
| *** achanda has quit IRC | 00:44 | |
| *** saikrishna has quit IRC | 00:52 | |
| *** shoutm has quit IRC | 01:01 | |
| *** vivekd has quit IRC | 01:04 | |
| *** vivekd has joined #openstack-nova | 01:05 | |
| *** shyu has joined #openstack-nova | 01:06 | |
| *** salv-orlando has joined #openstack-nova | 01:07 | |
| *** shoutm has joined #openstack-nova | 01:09 | |
| *** salv-orlando has quit IRC | 01:13 | |
| *** baoli has joined #openstack-nova | 01:13 | |
| *** salv-orlando has joined #openstack-nova | 01:13 | |
| *** sorrison has joined #openstack-nova | 01:17 | |
| *** baoli has quit IRC | 01:19 | |
| *** EinstCrazy has joined #openstack-nova | 01:19 | |
| *** zhangjn has joined #openstack-nova | 01:20 | |
| *** zhenguo has joined #openstack-nova | 01:24 | |
| openstackgerrit | Zhongcheng Lao proposed openstack/nova: VMware: Omit exceptions during snapshot cleanup https://review.openstack.org/216086 | 01:26 |
|---|---|---|
| *** salv-orlando has quit IRC | 01:26 | |
| *** vivekd has quit IRC | 01:32 | |
| *** baoli has joined #openstack-nova | 01:33 | |
| *** marzif__ has joined #openstack-nova | 01:33 | |
| *** tobasco_ has joined #openstack-nova | 01:36 | |
| *** tobasco has quit IRC | 01:37 | |
| *** marzif_ has quit IRC | 01:37 | |
| *** achanda has joined #openstack-nova | 01:37 | |
| *** hparekh has quit IRC | 01:38 | |
| *** hparekh2 has joined #openstack-nova | 01:38 | |
| *** takedakn has joined #openstack-nova | 01:39 | |
| *** hparekh has joined #openstack-nova | 01:40 | |
| *** zhangjn has quit IRC | 01:40 | |
| *** haomaiwang has joined #openstack-nova | 01:40 | |
| *** zhangjn has joined #openstack-nova | 01:41 | |
| *** liusheng has joined #openstack-nova | 01:41 | |
| *** aix_ has joined #openstack-nova | 01:41 | |
| *** vilobhmm1 has joined #openstack-nova | 01:42 | |
| *** hparekh2 has quit IRC | 01:42 | |
| *** aix has quit IRC | 01:42 | |
| *** shoutm has quit IRC | 01:42 | |
| *** shoutm has joined #openstack-nova | 01:42 | |
| *** vilobhmm11 has joined #openstack-nova | 01:43 | |
| *** baoli has quit IRC | 01:46 | |
| *** vilobhmm1 has quit IRC | 01:47 | |
| *** emagana has joined #openstack-nova | 01:56 | |
| alex_xu | oomichi: o ha yo! Does johnthetubaguy's reply address your concern https://review.openstack.org/#/c/214592/10 ? hope we can merge this in L3 | 01:57 |
| *** emagana_ has joined #openstack-nova | 01:58 | |
| *** baoli has joined #openstack-nova | 02:00 | |
| *** emagana has quit IRC | 02:01 | |
| *** lyanchih has joined #openstack-nova | 02:05 | |
| *** meghal1 has quit IRC | 02:07 | |
| oomichi | alex_xu: thanks for pointing it out. yeah, there are several patches around this(v3 -> v2.1). | 02:09 |
| oomichi | alex_xu: so I am fine for current patch now | 02:09 |
| alex_xu | oomichi: yea | 02:09 |
| alex_xu | oomichi: cool | 02:09 |
| alex_xu | oomichi: thanks | 02:09 |
| *** armax has joined #openstack-nova | 02:10 | |
| oomichi | alex_xu: np, done | 02:10 |
| *** takedakn has quit IRC | 02:12 | |
| openstackgerrit | Pengtao Huang proposed openstack/nova: Replace "all_mappings" variable by "block_device_mappings" https://review.openstack.org/208511 | 02:13 |
| *** flwang1 has quit IRC | 02:18 | |
| *** flwang1 has joined #openstack-nova | 02:19 | |
| *** zhangjn has quit IRC | 02:19 | |
| *** achanda has quit IRC | 02:22 | |
| *** achanda has joined #openstack-nova | 02:24 | |
| *** salv-orlando has joined #openstack-nova | 02:28 | |
| *** yuntongjin has quit IRC | 02:34 | |
| *** yuntongjin has joined #openstack-nova | 02:40 | |
| *** asettle has quit IRC | 02:40 | |
| *** yuntongjin1 has joined #openstack-nova | 02:41 | |
| *** salv-orlando has quit IRC | 02:41 | |
| *** asettle has joined #openstack-nova | 02:44 | |
| *** yuntongjin has quit IRC | 02:45 | |
| *** hakimo_ has joined #openstack-nova | 02:52 | |
| *** yuntongjin has joined #openstack-nova | 02:52 | |
| *** hakimo has quit IRC | 02:54 | |
| *** yuntongjin1 has quit IRC | 02:54 | |
| *** EinstCrazy has quit IRC | 02:57 | |
| *** haomaiwang has quit IRC | 03:03 | |
| *** haomaiw__ has joined #openstack-nova | 03:03 | |
| *** mcarden has joined #openstack-nova | 03:03 | |
| *** d34dh0r53 has quit IRC | 03:25 | |
| *** yixing has joined #openstack-nova | 03:25 | |
| *** jroll has quit IRC | 03:26 | |
| *** vilobhmm11 has quit IRC | 03:26 | |
| *** vilobhmm1 has joined #openstack-nova | 03:27 | |
| *** liusheng has quit IRC | 03:27 | |
| *** shyu has quit IRC | 03:28 | |
| *** liusheng has joined #openstack-nova | 03:28 | |
| *** apmelton has quit IRC | 03:28 | |
| *** shyu has joined #openstack-nova | 03:28 | |
| *** apmelton has joined #openstack-nova | 03:30 | |
| *** jroll has joined #openstack-nova | 03:31 | |
| *** d34dh0r53 has joined #openstack-nova | 03:32 | |
| *** I has joined #openstack-nova | 03:35 | |
| *** I is now known as Guest25059 | 03:36 | |
| *** links has joined #openstack-nova | 03:40 | |
| *** salv-orlando has joined #openstack-nova | 03:43 | |
| *** jwcroppe_ has joined #openstack-nova | 03:45 | |
| *** jwcroppe has quit IRC | 03:47 | |
| *** Guest25059 has quit IRC | 03:49 | |
| openstackgerrit | Eli Qiao proposed openstack/nova: API: Handle InstanceUnknownCell exceptions https://review.openstack.org/212965 | 03:49 |
| *** haigang has joined #openstack-nova | 03:49 | |
| *** vilobhmm1 has quit IRC | 03:50 | |
| *** vilobhmm1 has joined #openstack-nova | 03:51 | |
| openstackgerrit | lyanchih proposed openstack/nova: Creating instance fail when inject ssh key in cells mode https://review.openstack.org/197514 | 03:51 |
| *** sorrison has left #openstack-nova | 03:52 | |
| *** moshele has joined #openstack-nova | 03:53 | |
| *** takedakn has joined #openstack-nova | 03:53 | |
| *** Piet has joined #openstack-nova | 03:55 | |
| *** Piet has quit IRC | 03:56 | |
| *** salv-orlando has quit IRC | 03:56 | |
| *** deepthi has joined #openstack-nova | 03:56 | |
| *** Piet has joined #openstack-nova | 03:56 | |
| openstackgerrit | Zhongcheng Lao proposed openstack/nova: VMware: Omit exceptions during snapshot cleanup https://review.openstack.org/216086 | 03:59 |
| *** Piet has quit IRC | 03:59 | |
| *** yuntongjin has quit IRC | 03:59 | |
| *** lyanchih has quit IRC | 04:01 | |
| *** vilobhmm1 has quit IRC | 04:02 | |
| *** vilobhmm1 has joined #openstack-nova | 04:04 | |
| *** shoutm has quit IRC | 04:04 | |
| *** takedakn has quit IRC | 04:04 | |
| *** VW_ has quit IRC | 04:06 | |
| *** shoutm has joined #openstack-nova | 04:06 | |
| *** VW_ has joined #openstack-nova | 04:07 | |
| *** baoli has quit IRC | 04:07 | |
| *** moshele has quit IRC | 04:08 | |
| *** VW__ has joined #openstack-nova | 04:10 | |
| *** VW_ has quit IRC | 04:10 | |
| *** yixing1 has joined #openstack-nova | 04:14 | |
| *** yuntongjin has joined #openstack-nova | 04:15 | |
| *** VW__ has quit IRC | 04:16 | |
| *** yixing has quit IRC | 04:16 | |
| *** shyu has quit IRC | 04:26 | |
| *** zzzeek has joined #openstack-nova | 04:30 | |
| *** zzzeek has quit IRC | 04:31 | |
| *** garyk has joined #openstack-nova | 04:35 | |
| *** VikasC has joined #openstack-nova | 04:36 | |
| *** zhangjn has joined #openstack-nova | 04:37 | |
| *** takedakn has joined #openstack-nova | 04:47 | |
| *** deepakcs has joined #openstack-nova | 04:49 | |
| *** salv-orlando has joined #openstack-nova | 04:52 | |
| *** Sree has joined #openstack-nova | 04:53 | |
| *** Sree has quit IRC | 04:54 | |
| *** Sree has joined #openstack-nova | 04:55 | |
| *** Sukhdev has joined #openstack-nova | 04:56 | |
| *** salv-orlando has quit IRC | 04:57 | |
| *** yixing1 has left #openstack-nova | 04:57 | |
| *** salv-orlando has joined #openstack-nova | 04:57 | |
| *** moshele has joined #openstack-nova | 04:58 | |
| *** ig0r__ has joined #openstack-nova | 04:58 | |
| *** tojuvone has joined #openstack-nova | 04:58 | |
| *** ig0r_ has quit IRC | 04:59 | |
| *** yuntongjin has quit IRC | 05:00 | |
| *** VikasC has quit IRC | 05:03 | |
| *** Sree has quit IRC | 05:03 | |
| *** asettle has quit IRC | 05:06 | |
| *** asettle has joined #openstack-nova | 05:08 | |
| *** VikasC has joined #openstack-nova | 05:08 | |
| *** salv-orlando has quit IRC | 05:10 | |
| *** lyanchih has joined #openstack-nova | 05:11 | |
| *** takedakn has quit IRC | 05:11 | |
| *** asettle has quit IRC | 05:14 | |
| *** asettle has joined #openstack-nova | 05:14 | |
| *** takedakn has joined #openstack-nova | 05:18 | |
| *** Sree has joined #openstack-nova | 05:21 | |
| *** haigang has quit IRC | 05:23 | |
| openstackgerrit | garyk proposed openstack/nova: Compute: remove the 'method' variable from InstanceInvalidState https://review.openstack.org/218629 | 05:24 |
| openstackgerrit | garyk proposed openstack/nova: Compute: remove unused parameter https://review.openstack.org/218730 | 05:27 |
| *** oomichi has quit IRC | 05:27 | |
| *** dave-mccowan has quit IRC | 05:29 | |
| *** takedakn has quit IRC | 05:32 | |
| *** lyanchih has quit IRC | 05:34 | |
| *** baoli has joined #openstack-nova | 05:38 | |
| *** Murali has joined #openstack-nova | 05:39 | |
| *** eglynn__ has joined #openstack-nova | 05:41 | |
| *** baoli has quit IRC | 05:42 | |
| *** haigang has joined #openstack-nova | 05:47 | |
| *** markvoelker has joined #openstack-nova | 05:51 | |
| *** rotbeard has joined #openstack-nova | 05:53 | |
| *** markvoelker_ has joined #openstack-nova | 05:53 | |
| *** shoutm has quit IRC | 05:54 | |
| *** markvoelker has quit IRC | 05:56 | |
| *** sudipto has joined #openstack-nova | 06:00 | |
| *** shuquan has joined #openstack-nova | 06:05 | |
| *** ankit_ag has joined #openstack-nova | 06:06 | |
| *** lyanchih has joined #openstack-nova | 06:09 | |
| *** salv-orlando has joined #openstack-nova | 06:10 | |
| *** Sree_ has joined #openstack-nova | 06:14 | |
| *** Sree_ is now known as Guest41722 | 06:14 | |
| *** davideagnello has joined #openstack-nova | 06:16 | |
| *** vilobhmm1 has quit IRC | 06:16 | |
| *** Sree has quit IRC | 06:16 | |
| *** oomichi has joined #openstack-nova | 06:18 | |
| *** vilobhmm1 has joined #openstack-nova | 06:18 | |
| *** ishant has joined #openstack-nova | 06:18 | |
| *** eglynn__ has quit IRC | 06:18 | |
| *** afazekas_ has joined #openstack-nova | 06:19 | |
| *** eglynn__ has joined #openstack-nova | 06:20 | |
| *** markvoelker has joined #openstack-nova | 06:22 | |
| *** shuquan has quit IRC | 06:23 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Imported Translations from Transifex https://review.openstack.org/212367 | 06:23 |
| *** eglynn__ has quit IRC | 06:25 | |
| *** markvoelker_ has quit IRC | 06:25 | |
| *** shuquan has joined #openstack-nova | 06:26 | |
| *** markvoelker_ has joined #openstack-nova | 06:27 | |
| *** haigang has quit IRC | 06:27 | |
| *** armax has quit IRC | 06:28 | |
| *** haigang has joined #openstack-nova | 06:28 | |
| *** markvoelker has quit IRC | 06:29 | |
| *** sahid has joined #openstack-nova | 06:29 | |
| *** davideagnello has quit IRC | 06:29 | |
| *** haigang has quit IRC | 06:29 | |
| *** vilobhmm1 has quit IRC | 06:30 | |
| *** haigang has joined #openstack-nova | 06:36 | |
| *** VikasC has quit IRC | 06:40 | |
| *** shoutm has joined #openstack-nova | 06:43 | |
| *** zhangjn has quit IRC | 06:45 | |
| *** zhangjn has joined #openstack-nova | 06:45 | |
| openstackgerrit | garyk proposed openstack/nova: DB: add support for volume_type by BDMS https://review.openstack.org/139992 | 06:45 |
| openstackgerrit | garyk proposed openstack/nova: Object: add volume_type_id to block_device objects https://review.openstack.org/139994 | 06:45 |
| openstackgerrit | Hiroyuki Eguchi proposed openstack/nova: Cannot rebuild a instance booted from volume https://review.openstack.org/201458 | 06:50 |
| *** Guest41722 has quit IRC | 06:58 | |
| *** Sree has joined #openstack-nova | 06:59 | |
| *** shoutm_ has joined #openstack-nova | 06:59 | |
| *** Sukhdev has quit IRC | 07:00 | |
| *** shuquan has quit IRC | 07:00 | |
| *** shoutm has quit IRC | 07:00 | |
| *** eglynn__ has joined #openstack-nova | 07:02 | |
| *** stefan_amann has joined #openstack-nova | 07:03 | |
| *** Sree has quit IRC | 07:03 | |
| *** achanda has quit IRC | 07:05 | |
| *** eglynn__ has quit IRC | 07:09 | |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Switch to oslo.cache lib https://review.openstack.org/203049 | 07:20 |
| *** VikasC has joined #openstack-nova | 07:23 | |
| *** marcusvrn_ has quit IRC | 07:25 | |
| *** apuimedo_ has joined #openstack-nova | 07:27 | |
| *** romainh has joined #openstack-nova | 07:27 | |
| *** alexschm has joined #openstack-nova | 07:28 | |
| *** zhangjn has quit IRC | 07:28 | |
| *** asettle has quit IRC | 07:30 | |
| *** emagana_ has quit IRC | 07:31 | |
| *** emagana has joined #openstack-nova | 07:32 | |
| bauzas | johnthetubaguy: morning | 07:32 |
| bauzas | johnthetubaguy: when're you're there, ping me, I just want to disucss the gate stability, the cells job issue and the tagging happening tomorrow :) | 07:33 |
| *** kaisers has joined #openstack-nova | 07:33 | |
| *** doude has joined #openstack-nova | 07:34 | |
| openstackgerrit | Luo Gangyi proposed openstack/nova: Ensure bridge exists before resuming a VM after an offline snapshot https://review.openstack.org/149942 | 07:36 |
| *** emagana has quit IRC | 07:37 | |
| *** takedakn has joined #openstack-nova | 07:37 | |
| *** Sree has joined #openstack-nova | 07:37 | |
| *** yassine__ has joined #openstack-nova | 07:40 | |
| *** bfic has joined #openstack-nova | 07:44 | |
| *** takedakn has quit IRC | 07:47 | |
| *** markus_z has joined #openstack-nova | 07:47 | |
| *** ndipanov has joined #openstack-nova | 07:47 | |
| *** rdopiera has joined #openstack-nova | 07:47 | |
| *** takedakn has joined #openstack-nova | 07:47 | |
| *** takedakn has quit IRC | 07:49 | |
| *** Nikolay_St has joined #openstack-nova | 07:50 | |
| *** baoli has joined #openstack-nova | 07:58 | |
| garyk | bauzas: what tagging is tomorrow? | 08:00 |
| garyk | gate is borken | 08:00 |
| bauzas | garyk: l-3 and FF | 08:00 |
| garyk | ok, thanks. | 08:00 |
| alexschm | lyanchih: can you have a look at https://review.openstack.org/#/c/215102/ ? | 08:01 |
| alexschm | lyanchih: it should fix the problem you describe in bug 1489853 | 08:01 |
| openstack | bug 1489853 in OpenStack Compute (nova) "when hard-reboot a instance with serial-port mutil times ,instance will not start for port exhausted" [Undecided,New] https://launchpad.net/bugs/1489853 - Assigned to Alexander Schmidt (alexs-h) | 08:01 |
| lyanchih | alexschm: I will review that, thx. | 08:02 |
| alexschm | lyanchih: ok it's currently "work in progress" because it does not contain new unit tests, but the production code should be ready | 08:03 |
| *** baoli has quit IRC | 08:03 | |
| *** matrohon has joined #openstack-nova | 08:04 | |
| *** apuimedo_ has quit IRC | 08:05 | |
| *** ircuser-1 has joined #openstack-nova | 08:08 | |
| *** zhangjn has joined #openstack-nova | 08:10 | |
| *** claudiub has joined #openstack-nova | 08:10 | |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: Replace get_dynamic_property with get_object_property https://review.openstack.org/164220 | 08:12 |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: Replace get_dynamic_properties with get_object_properties_dict https://review.openstack.org/164604 | 08:12 |
| *** shuquan has joined #openstack-nova | 08:12 | |
| *** jistr has joined #openstack-nova | 08:15 | |
| *** kiran-r has joined #openstack-nova | 08:15 | |
| *** lucasagomes has joined #openstack-nova | 08:15 | |
| *** derekh has joined #openstack-nova | 08:15 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Hyper-V: Adds RDPConsoleOps unit tests https://review.openstack.org/139798 | 08:19 |
| *** lpetrut has joined #openstack-nova | 08:23 | |
| *** shuquan has quit IRC | 08:24 | |
| *** shuquan has joined #openstack-nova | 08:24 | |
| *** salv-orlando has quit IRC | 08:24 | |
| *** salv-orlando has joined #openstack-nova | 08:25 | |
| *** shuquan has quit IRC | 08:25 | |
| openstackgerrit | Sachi King proposed openstack/nova: Add constraint target to tox.ini https://review.openstack.org/205931 | 08:26 |
| *** shuquan has joined #openstack-nova | 08:26 | |
| johnthetubaguy | bauzas: hi | 08:27 |
| bauzas | johnthetubaguy: morning | 08:27 |
| johnthetubaguy | bauzas: I am actually away today till wednesday afternoon (yes crazy timing I know... but hence I am here right now) | 08:28 |
| bauzas | johnthetubaguy: oh ok | 08:28 |
| johnthetubaguy | bauzas: the gate doesn't look quite so bad this morning, 22 I see | 08:28 |
| johnthetubaguy | but I guess thats not the full story here | 08:28 |
| bauzas | johnthetubaguy: I'm mostly concerned by the failures | 08:28 |
| *** markvoelker_ has quit IRC | 08:28 | |
| bauzas | johnthetubaguy: https://bugs.launchpad.net/tempest/+bug/1489581 is still occuring | 08:29 |
| openstack | Launchpad bug 1489581 in OpenStack Compute (nova) "test_create_ebs_image_and_check_boot is race failing" [Low,Triaged] | 08:29 |
| bauzas | johnthetubaguy: even if the tempest bug has been delivered | 08:29 |
| *** lpetrut has quit IRC | 08:29 | |
| johnthetubaguy | so I guess we are running more jobs than ever, so we are hitting the races more than ever | 08:29 |
| bauzas | johnthetubaguy: but if you're off, don't worry, will speak to mriedem later | 08:29 |
| johnthetubaguy | or is this the newer one? | 08:29 |
| bauzas | johnthetubaguy: yeah, perhaps | 08:30 |
| johnthetubaguy | yeah, I mean would be awesome if someone who is around could help with that, would be awesome | 08:30 |
| johnthetubaguy | but honestly, its probably quite hard to land the bug fix | 08:30 |
| bauzas | johnthetubaguy: I'm already looking at a cells problem, can try to dig into that one | 08:30 |
| *** lpetrut has joined #openstack-nova | 08:30 | |
| bauzas | johnthetubaguy: agreed, will speak to alaski as well | 08:30 |
| johnthetubaguy | focus us good for these things | 08:30 |
| johnthetubaguy | you got a link to the cells one? | 08:31 |
| bauzas | johnthetubaguy: any idea when the tag is done ? | 08:31 |
| johnthetubaguy | bauzas: when we want, but by the latest on thursday | 08:31 |
| bauzas | johnthetubaguy: https://review.openstack.org/#/c/218473/1 | 08:31 |
| johnthetubaguy | bauzas: will be meeting with the release manager tomorrow to discuss that | 08:31 |
| bauzas | ack | 08:31 |
| johnthetubaguy | about this time tomorrow actually | 08:31 |
| bauzas | I'll leave that to ttx and you :à) | 08:31 |
| alex_xu | johnthetubaguy: morning, as you away these two days, can I update this patch https://review.openstack.org/#/c/214601/? and there is one comment from me, let me know your opionion, then I can update it. | 08:32 |
| bauzas | johnthetubaguy: take good rest | 08:32 |
| bauzas | johnthetubaguy: can handle all of that | 08:32 |
| *** aix_ has quit IRC | 08:32 | |
| johnthetubaguy | alex_xu: free free to update any of my patches, sorry, forgot to reach out to you on Friday | 08:33 |
| *** aix has joined #openstack-nova | 08:33 | |
| *** eglynn__ has joined #openstack-nova | 08:33 | |
| johnthetubaguy | alex_xu: I added comments about M though | 08:33 |
| *** emagana has joined #openstack-nova | 08:33 | |
| alex_xu | johnthetubaguy: don't worry, let me take care the patches | 08:33 |
| johnthetubaguy | alex_xu: sdague asked my to remove M from most of them, he had a good point, we are not sure when we will remove it yet | 08:33 |
| johnthetubaguy | alex_xu: they all had M in the first place, but changed my mind | 08:33 |
| alex_xu | johnthetubaguy: ok, got it | 08:34 |
| johnthetubaguy | alex_xu: I am expecting operator push back, which will guide the timeframe, basically | 08:34 |
| alex_xu | johnthetubaguy: ok, so let me remove M from the message | 08:34 |
| johnthetubaguy | alex_xu: where does it say M? | 08:35 |
| johnthetubaguy | oh, I see now | 08:35 |
| johnthetubaguy | oops | 08:35 |
| alex_xu | :) | 08:36 |
| *** shoutm_ has quit IRC | 08:36 | |
| johnthetubaguy | alex_xu: I added an extra comment, there are two Ms in that file, you probably saw that already | 08:37 |
| alex_xu | johnthetubaguy: ok, thanks | 08:37 |
| *** emagana has quit IRC | 08:38 | |
| *** shuquan has quit IRC | 08:38 | |
| johnthetubaguy | alex_xu: how are we lucking for all the API patches now? | 08:39 |
| johnthetubaguy | alex_xu: those renames we can merge after the freeze, I feel | 08:39 |
| alex_xu | johnthetubaguy: yea, rename isn't hurry for now | 08:39 |
| alex_xu | johnthetubaguy: https://etherpad.openstack.org/p/liberty-nova-priorities-tracking | 08:39 |
| alex_xu | johnthetubaguy: just left 2 patches for L3 | 08:40 |
| johnthetubaguy | I am seeing +W on the two patches that really worry me | 08:40 |
| alex_xu | johnthetubaguy: which one worry you? | 08:40 |
| *** amotoki has joined #openstack-nova | 08:41 | |
| openstackgerrit | Sachi King proposed openstack/nova: Add constraint target to tox.ini https://review.openstack.org/205931 | 08:41 |
| johnthetubaguy | its the deprecate v3 and the api-paste.ini | 08:41 |
| *** dguitarbite has joined #openstack-nova | 08:42 | |
| *** haigang has quit IRC | 08:44 | |
| alex_xu | johnthetubaguy: anything you worry about? | 08:44 |
| *** haigang has joined #openstack-nova | 08:44 | |
| johnthetubaguy | alex_xu: lots of things, I can be a worry a lot at times, but I think we are looking good | 08:45 |
| johnthetubaguy | alex_xu: just spotted this patch: https://review.openstack.org/#/c/218103/1 | 08:45 |
| johnthetubaguy | alex_xu: looks good, added a comment about the v3 router, I think we have to keep that for one more release :( | 08:45 |
| *** zhangjn has quit IRC | 08:45 | |
| johnthetubaguy | only just dropped it out of the default api-paste.ini, oops | 08:45 |
| *** moshele has quit IRC | 08:46 | |
| alex_xu | johnthetubaguy: ok, got it, it's fine now | 08:46 |
| openstackgerrit | Zhenzan Zhou proposed openstack/nova: filter pre_assigned_dev_names when finding disk dev https://review.openstack.org/214419 | 08:46 |
| johnthetubaguy | ndipanov: it would be awesome if you can take a peak at this cells bdm thing please: https://review.openstack.org/#/c/218450/ | 08:49 |
| ndipanov | johnthetubaguy, will do - this is what I was chatting to alaski on Fri about | 08:49 |
| johnthetubaguy | ndipanov: cool | 08:50 |
| *** haigang has quit IRC | 08:50 | |
| ndipanov | johnthetubaguy, a very easy one back at you https://review.openstack.org/#/c/218201/ | 08:50 |
| ndipanov | but don't stop there :) | 08:51 |
| bauzas | ndipanov: yeah that's what I was discussing with johnthetubaguy | 08:51 |
| *** VikasC has quit IRC | 08:51 | |
| ndipanov | bauzas, which one? | 08:51 |
| bauzas | ndipanov: I think alaski's change is good because it could potentially not hurt https://bugs.launchpad.net/tempest/+bug/1489581 | 08:51 |
| openstack | Launchpad bug 1489581 in OpenStack Compute (nova) "test_create_ebs_image_and_check_boot is race failing" [Low,Triaged] | 08:51 |
| johnthetubaguy | ndipanov: ah, didn't see your comment, got it now | 08:52 |
| bauzas | ndipanov: which is totally boring for passing the gate | 08:52 |
| *** haigang has joined #openstack-nova | 08:52 | |
| openstackgerrit | Alex Xu proposed openstack/nova: api: deprecate the api v2 extension configuration https://review.openstack.org/214601 | 08:53 |
| openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova: libvirt: make live_migration_uri flag dependent on virt_type https://review.openstack.org/175780 | 08:54 |
| *** jichen has joined #openstack-nova | 08:56 | |
| *** MarcF has joined #openstack-nova | 08:57 | |
| *** bfic has quit IRC | 08:58 | |
| *** bfic has joined #openstack-nova | 09:00 | |
| *** ildikov has joined #openstack-nova | 09:02 | |
| lyanchih | alexschm: The patch looks can fix 1489853. | 09:03 |
| alexschm | lyanchih: yes I think so too, it will return the already acquired port when get_guest_xml is called multiple times | 09:04 |
| lyanchih | But why you don’t using singleton class such as SerialPortPool, which can packet those variable like ALLOCATED_PORTS or SERIAL_PORTS | 09:04 |
| *** MarcF has quit IRC | 09:05 | |
| lyanchih | It looks also work if users change start number. | 09:07 |
| *** IanGovett has joined #openstack-nova | 09:08 | |
| lyanchih | BTW, it can also fix https://bugs.launchpad.net/nova/+bug/1480514 | 09:08 |
| openstack | Launchpad bug 1480514 in OpenStack Compute (nova) "Remove error instance fail when enable serial_consol" [Undecided,In progress] - Assigned to Chung Chih, Hung (lyanchih) | 09:08 |
| alexschm | lyanchih: you mean when a user changes serial console port_range? | 09:10 |
| lyanchih | yes | 09:10 |
| *** moshele has joined #openstack-nova | 09:10 | |
| gmann | jichen: ping | 09:10 |
| alexschm | I think for this configuration change to become effective, restarting nova compute is required | 09:11 |
| jichen | gmann: ? | 09:11 |
| gmann | jichen: can you please check my reply on https://review.openstack.org/#/c/217572/ | 09:11 |
| jichen | gmann: OK, in a few minutes | 09:11 |
| gmann | jichen: if that looks ok to you | 09:11 |
| gmann | jichen: Thanks :) | 09:11 |
| jichen | gmann: np :) | 09:11 |
| alexschm | lyanchih: but using a singleton class sounds like a good idea, maybe you can suggest this via a code review | 09:12 |
| gmann | johnthetubaguy: please look on VIF thing back-porting to kilo - https://review.openstack.org/#/c/198944/ | 09:13 |
| openstackgerrit | garyk proposed openstack/nova: VMware: adds support for rescue image https://review.openstack.org/218776 | 09:15 |
| lyanchih | alexschm: Thx, I will comment it at code review. | 09:15 |
| gmann | johnthetubaguy: oomichi : alex_xu : this too if we can make it - https://review.openstack.org/#/c/217572/ | 09:15 |
| lyanchih | alexschm: Bug 1480514 will also been fixed. Do I need to do in launch pad? | 09:15 |
| openstack | bug 1480514 in OpenStack Compute (nova) "Remove error instance fail when enable serial_consol" [Undecided,In progress] https://launchpad.net/bugs/1480514 - Assigned to Chung Chih, Hung (lyanchih) | 09:15 |
| alexschm | lyanchih: I'll also assign it to myself and mention it in patchset 215102 if that is okay with you? | 09:17 |
| lyanchih | That’s ok. I will also abandon my review. | 09:18 |
| alexschm | lyanchih: but the bugfix suggested for 1480514 also looks good by itself, so we'll simply see what gets merged first | 09:18 |
| *** MarcF has joined #openstack-nova | 09:19 | |
| alexschm | lyanchih: I suggest you keep it, as it's small and will probably get merged faster than mine | 09:19 |
| *** haigang has quit IRC | 09:20 | |
| *** zhangjn has joined #openstack-nova | 09:23 | |
| *** markvoelker has joined #openstack-nova | 09:23 | |
| lyanchih | alexschm: Ok, thx. But it didn’t have news recently. :( | 09:24 |
| *** salv-orlando has quit IRC | 09:24 | |
| alexschm | lyanchih: it just got another +1 | 09:25 |
| *** markvoelker_ has joined #openstack-nova | 09:26 | |
| *** zhangjn has quit IRC | 09:27 | |
| lyanchih | Ha, thank you. | 09:27 |
| *** bfic has quit IRC | 09:29 | |
| *** markvoelker has quit IRC | 09:30 | |
| *** markvoelker_ has quit IRC | 09:31 | |
| *** alex_klimov has joined #openstack-nova | 09:31 | |
| *** haigang has joined #openstack-nova | 09:33 | |
| openstackgerrit | Sudipta Biswas proposed openstack/nova: Allow monitor plugins to set own object fields https://review.openstack.org/214772 | 09:35 |
| *** ig0r_ has joined #openstack-nova | 09:37 | |
| *** markvoelker has joined #openstack-nova | 09:37 | |
| *** marzif__ has quit IRC | 09:38 | |
| *** marzif__ has joined #openstack-nova | 09:38 | |
| *** rdopiera has quit IRC | 09:38 | |
| *** ig0r__ has quit IRC | 09:39 | |
| *** zhangjn has joined #openstack-nova | 09:39 | |
| *** zhangjn has quit IRC | 09:41 | |
| *** VikasC has joined #openstack-nova | 09:47 | |
| ndipanov | https://review.openstack.org/#/c/218450/ johnthetubaguy done | 09:48 |
| ndipanov | should I ack it ? | 09:48 |
| ndipanov | or? | 09:48 |
| *** aix has quit IRC | 09:49 | |
| *** amann has joined #openstack-nova | 09:49 | |
| *** amann has quit IRC | 09:49 | |
| *** s_amann has joined #openstack-nova | 09:49 | |
| ndipanov | done | 09:50 |
| *** romainh has quit IRC | 09:50 | |
| *** stefan_amann has quit IRC | 09:52 | |
| *** vivekd has joined #openstack-nova | 09:54 | |
| *** Sree__ has joined #openstack-nova | 09:58 | |
| openstackgerrit | sahid proposed openstack/nova: libvirt: make detach_device to return async object https://review.openstack.org/217680 | 09:59 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: Force iSCSI disconnect after timeout https://review.openstack.org/167815 | 09:59 |
| *** yassine__ has quit IRC | 10:00 | |
| *** Sree has quit IRC | 10:01 | |
| *** moshele has quit IRC | 10:08 | |
| *** takedakn has joined #openstack-nova | 10:11 | |
| *** marzif__ has quit IRC | 10:12 | |
| *** takedakn has quit IRC | 10:14 | |
| *** haigang has quit IRC | 10:14 | |
| *** takedakn has joined #openstack-nova | 10:14 | |
| *** haigang has joined #openstack-nova | 10:15 | |
| *** moshele has joined #openstack-nova | 10:15 | |
| *** takedakn has quit IRC | 10:17 | |
| *** takedakn has joined #openstack-nova | 10:18 | |
| *** aix has joined #openstack-nova | 10:18 | |
| *** MarcF has quit IRC | 10:21 | |
| *** Sree__ has quit IRC | 10:23 | |
| *** ferest has joined #openstack-nova | 10:25 | |
| *** shoutm has joined #openstack-nova | 10:27 | |
| *** ferest has quit IRC | 10:29 | |
| *** MarcF has joined #openstack-nova | 10:34 | |
| *** lyanchih has quit IRC | 10:38 | |
| openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova: libvirt: make live_migration_uri flag dependent on virt_type https://review.openstack.org/175780 | 10:39 |
| *** pixelb has joined #openstack-nova | 10:43 | |
| openstackgerrit | sulochan-acharya proposed openstack/nova: xapi: cleanup volume sr on live migration rollback https://review.openstack.org/217105 | 10:44 |
| openstackgerrit | garyk proposed openstack/nova: DB: add support for volume_type by BDMS https://review.openstack.org/139992 | 10:45 |
| openstackgerrit | garyk proposed openstack/nova: Object: add volume_type_id to block_device objects https://review.openstack.org/139994 | 10:45 |
| *** marzif__ has joined #openstack-nova | 10:47 | |
| *** shoutm has quit IRC | 10:56 | |
| *** jaypipes has joined #openstack-nova | 10:58 | |
| *** gszasz has joined #openstack-nova | 11:01 | |
| *** takedakn has quit IRC | 11:01 | |
| *** qianli has joined #openstack-nova | 11:01 | |
| openstackgerrit | Alex Xu proposed openstack/nova: api: deprecate the api v2 extension configuration https://review.openstack.org/214601 | 11:02 |
| *** dguitarbite has quit IRC | 11:04 | |
| *** yassine__ has joined #openstack-nova | 11:04 | |
| *** yassine__ has quit IRC | 11:04 | |
| *** tojuvone has quit IRC | 11:10 | |
| jaypipes | morning all | 11:11 |
| *** takedakn has joined #openstack-nova | 11:11 | |
| jaypipes | bauzas, alaski: if you have a chance, this patch series is review-ready: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:pci-cleanup,n,z. And https://review.openstack.org/209499 is also ready. | 11:12 |
| *** takedakn has quit IRC | 11:13 | |
| jaypipes | bauzas: see my and dansmith's note on https://review.openstack.org/#/c/213591/? | 11:13 |
| *** takedakn has joined #openstack-nova | 11:13 | |
| openstackgerrit | Pranali Deore proposed openstack/nova: Fix order of arguments in assertEqual https://review.openstack.org/218809 | 11:14 |
| *** haomaiw__ has quit IRC | 11:15 | |
| *** shoutm has joined #openstack-nova | 11:19 | |
| jaypipes | alaski, mikal, Vek, _cerberus_: easy review bait for a couple small bug fixes from snikitin: https://review.openstack.org/#/c/214189/ and https://review.openstack.org/#/c/207832/ | 11:19 |
| *** yassine__ has joined #openstack-nova | 11:20 | |
| *** dguitarbite has joined #openstack-nova | 11:20 | |
| *** deepthi has quit IRC | 11:22 | |
| openstackgerrit | Pranali Deore proposed openstack/nova: Fix order of arguments in assertEqual https://review.openstack.org/218811 | 11:25 |
| *** ishant has quit IRC | 11:25 | |
| *** jcoufal has joined #openstack-nova | 11:28 | |
| *** rotbeard has quit IRC | 11:31 | |
| openstackgerrit | Pranali Deore proposed openstack/nova: Fix order of arguments in assertEqual https://review.openstack.org/218813 | 11:31 |
| *** lucasagomes is now known as lucas-hungry | 11:32 | |
| *** emagana has joined #openstack-nova | 11:32 | |
| *** jaypipes has quit IRC | 11:35 | |
| *** hparekh has quit IRC | 11:35 | |
| *** hparekh has joined #openstack-nova | 11:35 | |
| *** jaypipes has joined #openstack-nova | 11:36 | |
| *** emagana has quit IRC | 11:37 | |
| *** amotoki has quit IRC | 11:39 | |
| *** bfic has joined #openstack-nova | 11:39 | |
| *** amotoki has joined #openstack-nova | 11:39 | |
| *** marzif__ has quit IRC | 11:42 | |
| openstackgerrit | garyk proposed openstack/nova: Compute: remove unused parameter https://review.openstack.org/218730 | 11:43 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova: Add functional test for server group https://review.openstack.org/147514 | 11:43 |
| *** openstackgerrit has quit IRC | 11:46 | |
| *** openstackgerrit has joined #openstack-nova | 11:47 | |
| garyk | johnthetubaguy: what do you suggest that we do about https://review.openstack.org/190047 | 11:49 |
| *** ftersin has quit IRC | 11:49 | |
| garyk | i wrote a mail to the list and i felt that none of the people who responded to that could provide a good argument to keep this blocked. | 11:49 |
| garyk | it is a real issue that we have in deployment. | 11:50 |
| *** sudipto has quit IRC | 11:50 | |
| *** markvoelker_ has joined #openstack-nova | 11:51 | |
| *** markvoelker has quit IRC | 11:51 | |
| openstackgerrit | Pranali Deore proposed openstack/nova: Fix order of arguments in assertEqual https://review.openstack.org/218817 | 11:54 |
| *** markvoelker_ has quit IRC | 11:55 | |
| *** markvoelker has joined #openstack-nova | 11:56 | |
| *** mwagner_lap has quit IRC | 11:59 | |
| *** tojuvone has joined #openstack-nova | 12:01 | |
| *** cristicalin has joined #openstack-nova | 12:03 | |
| *** IanGovett1 has joined #openstack-nova | 12:05 | |
| *** IanGovett has quit IRC | 12:05 | |
| *** vivekd has quit IRC | 12:07 | |
| *** dprince has joined #openstack-nova | 12:08 | |
| *** dave-mccowan has joined #openstack-nova | 12:08 | |
| *** romainh has joined #openstack-nova | 12:09 | |
| *** aysyd has joined #openstack-nova | 12:14 | |
| *** raildo-afk is now known as raildo | 12:16 | |
| *** rfolco has joined #openstack-nova | 12:19 | |
| *** hparekh2 has joined #openstack-nova | 12:23 | |
| *** rdopiera has joined #openstack-nova | 12:24 | |
| *** hparekh has quit IRC | 12:24 | |
| *** ferest has joined #openstack-nova | 12:26 | |
| *** mhorban has quit IRC | 12:26 | |
| *** humble__ has quit IRC | 12:27 | |
| *** baoli has joined #openstack-nova | 12:29 | |
| bauzas | jaypipes: morning you | 12:30 |
| bauzas | jaypipes: sure, it was in my today's list | 12:30 |
| bauzas | jaypipes: I'm currently looking at the gate status | 12:30 |
| *** baoli_ has joined #openstack-nova | 12:31 | |
| *** karimb has joined #openstack-nova | 12:31 | |
| *** baoli_ has quit IRC | 12:31 | |
| bauzas | jaypipes: also, wanted to chat with you about https://review.openstack.org/#/c/215471/ | 12:32 |
| *** edmondsw has joined #openstack-nova | 12:32 | |
| *** baoli_ has joined #openstack-nova | 12:32 | |
| *** yuntongjin has joined #openstack-nova | 12:33 | |
| *** baoli has quit IRC | 12:34 | |
| *** ferest has quit IRC | 12:34 | |
| *** krtaylor has quit IRC | 12:34 | |
| *** Sree has joined #openstack-nova | 12:35 | |
| *** mhorban has joined #openstack-nova | 12:36 | |
| *** VikasC has quit IRC | 12:37 | |
| *** haomaiwa_ has joined #openstack-nova | 12:37 | |
| *** takedakn has quit IRC | 12:41 | |
| *** edleafe has joined #openstack-nova | 12:41 | |
| *** takedakn has joined #openstack-nova | 12:42 | |
| alex_xu | bauzas: jaypipes morning, looks like this break the server group https://review.openstack.org/#/c/197912/24 | 12:42 |
| alex_xu | which I reviewed at the weekend | 12:42 |
| bauzas | alex_xu: lemme doublecheck | 12:43 |
| alex_xu | bauzas: let me know if I missing something | 12:43 |
| bauzas | alex_xu: so the filt_props dict was having a 'group_hosts' keyt | 12:44 |
| bauzas | alex_xu: which is a list of hosts | 12:44 |
| *** annegentle has joined #openstack-nova | 12:44 | |
| bauzas | alex_xu: on the other hand, instance_group is an object | 12:44 |
| alex_xu | bauzas: yea, instance_group.members is a list of instance_uuid | 12:45 |
| bauzas | alex_xu: having a list of members | 12:45 |
| bauzas | alex_xu: okay, we need to check the filter then | 12:46 |
| bauzas | alex_xu: or where it's called | 12:46 |
| *** krtaylor has joined #openstack-nova | 12:47 | |
| alex_xu | bauzas: here https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/models.py#L1360 | 12:48 |
| *** vladikr has joined #openstack-nova | 12:48 | |
| bauzas | alex_xu: well, it works because it's a new object that I'm instanciating but not saving, right. | 12:49 |
| bauzas | ? | 12:49 |
| bauzas | alex_xu: but yeah that's a misusage of instance_group | 12:49 |
| *** burgerk has joined #openstack-nova | 12:50 | |
| jaypipes | bauzas: technically, it should be a set() of hosts. | 12:50 |
| *** lucas-hungry is now known as lucasagomes | 12:50 | |
| bauzas | jaypipes: yeah, like I did for the Retry object | 12:50 |
| *** MarcF has quit IRC | 12:50 | |
| bauzas | 'members' | 12:51 |
| jaypipes | right. | 12:51 |
| bauzas | what a nice name | 12:51 |
| bauzas | okay, I can do the change | 12:51 |
| jaypipes | bauzas: ty sir. alex_xu nice catch! | 12:51 |
| bauzas | but since the object is now merged, that means a version bump... | 12:51 |
| alex_xu | jaypipes: thanks | 12:51 |
| alex_xu | bauzas: change the instance group object? | 12:52 |
| bauzas | alex_xu: no, deprecate that field and provide a new field for that | 12:52 |
| bauzas | alex_xu: I mean the "instance_group" field | 12:52 |
| alex_xu | bauzas: emm...a little late at this point | 12:53 |
| bauzas | alex_xu: zactly | 12:53 |
| bauzas | alex_xu: that's the problem | 12:53 |
| alex_xu | bauzas: or leave this part in M? we can merge all the other part | 12:53 |
| alex_xu | bauzas: then we can fix that early when M open | 12:54 |
| bauzas | alex_xu: like I said, it works because it's not used as a persistent object | 12:54 |
| alex_xu | bauzas: emm....yea, but I'm not sure that is ok | 12:55 |
| bauzas | jaypipes: thoughts ? | 12:55 |
| bauzas | jaypipes: looks like you're an o.vo guru :p | 12:55 |
| bauzas | and superdan is not yet awaken :p | 12:55 |
| jaypipes | bauzas: removed my -1 and added +W on https://review.openstack.org/#/c/215471/8, but it needs a rebase and conflict resolve. | 12:55 |
| bauzas | jaypipes: coolness | 12:56 |
| *** afazekas__ has joined #openstack-nova | 12:56 | |
| bauzas | jaypipes: so, yeah, given all the issues we have with uncorrect fields, I know it really depends on the situation | 12:56 |
| bauzas | jaypipes: if I wanted to do that properly, it would require to create a new object which would be quite similar to the InstanceGroup object (ie. having policies and members) but where members would be a ListOfComputes | 12:58 |
| bauzas | jaypipes: and bump the versions etc. | 12:58 |
| *** afazekas_ has quit IRC | 12:58 | |
| bauzas | alex_xu: https://github.com/openstack/nova/blob/master/nova/objects/request_spec.py#L142 was merged :( | 12:59 |
| alex_xu | bauzas: yea :( | 13:00 |
| *** yuntongjin has quit IRC | 13:02 | |
| edleafe | good UGT morning, all | 13:02 |
| alex_xu | bauzas: looks like the correct way is we should report the instance group info into the scheduler. | 13:03 |
| alex_xu | bauzas: but this shouldn't happened in this point | 13:03 |
| bauzas | alex_xu: that's what we already do | 13:04 |
| bauzas | alex_xu: we're passing some dict to the schedulker | 13:04 |
| *** edmondsw has quit IRC | 13:04 | |
| bauzas | alex_xu: and the scheduler consumes it | 13:04 |
| *** dustins has joined #openstack-nova | 13:04 | |
| *** mjturek1 has joined #openstack-nova | 13:05 | |
| *** MarcF has joined #openstack-nova | 13:05 | |
| *** vivekd has joined #openstack-nova | 13:06 | |
| bauzas | jaypipes: quick q, can we have some fields in a NovaObject that can be persisted and some others not ? I know it's pretty bad tho | 13:07 |
| alex_xu | bauzas: emm...I need thnk about more | 13:07 |
| bauzas | jaypipes: I don't want to persist the list of hosts in the InstanceGroup object hence why | 13:07 |
| *** shuquan has joined #openstack-nova | 13:11 | |
| bauzas | superdan: ^ when you'll be caffeinated, I'd enjoy some discussion about that | 13:12 |
| jaypipes | bauzas: I'm looking at it again... | 13:12 |
| bauzas | in theory, the Facade is the object and we are not contracting the fact that the SQLA model and the object model are identical | 13:12 |
| jaypipes | bauzas: so, even if alex_xu is correct, this looks like it may be a bug in nova that precedes your patch, no? | 13:13 |
| bauzas | jaypipes: the bug is that the field is badly named | 13:13 |
| bauzas | and there is also a PEBKAC | 13:13 |
| bauzas | because I haven't fully thought about that | 13:13 |
| markus_z | anthonyper: Are you around? | 13:13 |
| bauzas | jaypipes: but since it's now merged for the eternity, we have a few options | 13:14 |
| bauzas | jaypipes: and I really think we should keep a nested InstanceGroup object in the ReqSpec object | 13:14 |
| bauzas | jaypipes: so my proposal would be : just dump a new version of InstanceGroup which would add a new field called "hosts" but not mapped with the SQLA model | 13:15 |
| jaypipes | bauzas: there is already a @remoteable method called get_hosts() on the InstanceGroup object... | 13:15 |
| *** xyang1 has joined #openstack-nova | 13:16 | |
| jaypipes | bauzas: I think we should add a new field to RequestSpec which is a DictOfSetsField() called instance_group_hosts. make it non-persisted. | 13:16 |
| jaypipes | alex_xu: AGREE? ^ | 13:17 |
| jaypipes | oops, sorry for yelling... | 13:17 |
| jaypipes | capslock fail | 13:17 |
| bauzas | jaypipes: I'd rather prefer to add the hosts field in the InstanceGroup object because of the semantics | 13:17 |
| bauzas | jaypipes: even if we already have a remotable for thzat | 13:17 |
| *** jcoufal has quit IRC | 13:17 | |
| bauzas | jaypipes: that looks even nicier : | 13:17 |
| *** mdrabe has joined #openstack-nova | 13:17 | |
| jaypipes | bauzas: and how would the hosts field get initially populated? | 13:18 |
| bauzas | jaypipes: something list obj.hosts = obj.get_hosts() | 13:18 |
| bauzas | jaypipes: thanks to what's already merged | 13:18 |
| bauzas | jaypipes: https://github.com/openstack/nova/blob/master/nova/objects/request_spec.py#L142 | 13:18 |
| *** ankit_ag has quit IRC | 13:19 | |
| jaypipes | bauzas: I think I would prefer to use expected_attrs='hosts' or something like that in the constructor. | 13:19 |
| *** kbyrne has quit IRC | 13:19 | |
| bauzas | jaypipes: you mean in _from_db_object() ? | 13:20 |
| jaypipes | bauzas: yeah, sorry.. | 13:21 |
| alex_xu | jaypipes: hosts field in InstanceGroup sounds better | 13:21 |
| jaypipes | alex_xu: yeah, cleaner... agreed. | 13:21 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add cpu_allocation_ratio and ram_allocation_ratio to ComputeNode https://review.openstack.org/215471 | 13:21 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update HostManager and filters to use ComputeNode ratios https://review.openstack.org/216696 | 13:21 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update ComputeNode values with allocation ratios in the RT https://review.openstack.org/216362 | 13:21 |
| bauzas | jaypipes: ^ any +W could be just awesomenessable | 13:22 |
| *** ivasev has joined #openstack-nova | 13:23 | |
| *** ssurana has joined #openstack-nova | 13:24 | |
| alex_xu | jaypipes: but that sounds like we need db call in the beginning of invoking scheduler | 13:24 |
| bauzas | alex_xu: jaypipes: nope we don't | 13:25 |
| *** mc has joined #openstack-nova | 13:25 | |
| bauzas | alex_xu: jaypipes: again, see https://github.com/openstack/nova/blob/master/nova/objects/request_spec.py#L133-L144 | 13:26 |
| alex_xu | jaypipes: bauzas as line 67 https://review.openstack.org/#/c/188487/30/nova/scheduler/filter_scheduler.py we build request_spec obj in the beginning | 13:26 |
| *** mc is now known as Guest15619 | 13:26 | |
| openstackgerrit | Pengtao Huang proposed openstack/nova: Replace "all_mappings" variable by "block_device_mappings" https://review.openstack.org/208511 | 13:26 |
| bauzas | alex_xu: if you read my series, you see that I'm using some helpers for hydrating the object | 13:26 |
| alex_xu | bauzas: ah, I see, that sounds better | 13:26 |
| *** zzzeek has joined #openstack-nova | 13:27 | |
| *** nkrinner has quit IRC | 13:27 | |
| bauzas | alex_xu: so, in Mitaka, I'm planning to completely remove the build_req_spec() helper and directly hydrate the Spec object in the conductor | 13:27 |
| bauzas | okay, so here the deal | 13:27 |
| jaypipes | bauzas: done (allocation series +W'd) | 13:27 |
| bauzas | jaypipes: Supercalifragilisticexpialidocious | 13:27 |
| *** eharney has joined #openstack-nova | 13:27 | |
| *** ijuwang has joined #openstack-nova | 13:28 | |
| * jaypipes wonders how alex_xu would translate "Supercalifragilisticexpialidocious" to Mandarin.. :) | 13:28 | |
| * alex_xu really have no idea what is that... | 13:28 | |
| *** liusheng has quit IRC | 13:28 | |
| jaypipes | alex_xu: https://en.wikipedia.org/wiki/Supercalifragilisticexpialidocious | 13:28 |
| jaypipes | :) | 13:29 |
| *** mcarden has quit IRC | 13:29 | |
| *** mtanino has joined #openstack-nova | 13:30 | |
| *** bfic has left #openstack-nova | 13:30 | |
| *** links has quit IRC | 13:30 | |
| *** VW_ has joined #openstack-nova | 13:30 | |
| *** jcoufal has joined #openstack-nova | 13:30 | |
| bauzas | alex_xu: jaypipes: okay, lemme draw something for you guys (re: the spec object) | 13:30 |
| bauzas | basically, bumping a new field 'hosts' and populate it correctly in _from_db_objects() | 13:31 |
| bauzas | with an expected_attrs and a lazy-load | 13:31 |
| *** jcoufal has quit IRC | 13:31 | |
| jaypipes | bauzas: yes. | 13:32 |
| *** kbyrne has joined #openstack-nova | 13:32 | |
| bauzas | not sure for the lazy-load, the idea is that if the field is unset, we lazy-load thanks to get_hosts() | 13:32 |
| alex_xu | jaypipes: ok, Supercalifragilisticexpialidocious means 'agree'? | 13:32 |
| bauzas | alex_xu: means it's a mix of feelings | 13:32 |
| *** thorst1 has joined #openstack-nova | 13:32 | |
| *** jcoufal has joined #openstack-nova | 13:32 | |
| bauzas | alex_xu: but basically you're happy | 13:32 |
| *** edmondsw has joined #openstack-nova | 13:33 | |
| alex_xu | bauzas: heh | 13:33 |
| bauzas | The roots of the word have been defined[4] as follows: super- "above", cali- "beauty", fragilistic- "delicate", expiali- "to atone", and -docious "educable", with the sum of these parts signifying roughly "Atoning for educability through delicate beauty." According to the film, it is defined as "something to say when you have nothing to say". | 13:33 |
| bauzas | not sure it's an official word tho :) | 13:33 |
| bauzas | I just have 2 daughters with one by 5yo | 13:34 |
| *** yuntongjin has joined #openstack-nova | 13:34 | |
| * alex_xu try to find if there someting similar to Mandarin... | 13:34 | |
| *** burgerk has quit IRC | 13:34 | |
| *** IanGovett1 has quit IRC | 13:35 | |
| superdan | jaypipes: heh | 13:35 |
| *** superdan is now known as dansmith | 13:35 | |
| dansmith | bauzas: discussion about what? persisting only a few fields of an object? | 13:35 |
| bauzas | dansmith: already awake? woah | 13:36 |
| *** salv-orlando has joined #openstack-nova | 13:36 | |
| *** dims has joined #openstack-nova | 13:36 | |
| dansmith | bauzas: same time every day... | 13:36 |
| bauzas | dansmith: so, there was a PEBKAC here | 13:36 |
| bauzas | dansmith: and alex_xu just found it | 13:36 |
| *** rushil has joined #openstack-nova | 13:36 | |
| bauzas | dansmith: https://review.openstack.org/#/c/197912/24/nova/scheduler/filter_scheduler.py,cm | 13:37 |
| bauzas | voilà | 13:37 |
| bauzas | dansmith: so, my proposal was | 13:37 |
| *** rushil_ has joined #openstack-nova | 13:37 | |
| bauzas | (15:31:24) bauzas: basically, bumping a new field 'hosts' and populate it correctly in _from_db_objects() | 13:38 |
| bauzas | (15:31:35) bauzas: with an expected_attrs and a lazy-load | 13:38 |
| bauzas | (15:32:07) bauzas: not sure for the lazy-load, the idea is that if the field is unset, we lazy-load thanks to get_hosts() | 13:38 |
| bauzas | dansmith: so, no need to change the spec object, just bumping a new version | 13:38 |
| bauzas | dansmith: my thoughts was going to having DB and Object models which were different, but I answered my question | 13:38 |
| *** thangp has joined #openstack-nova | 13:39 | |
| dansmith | okay, not sure I understand what you're saying, but having a (real) lazy-load in the scheduler would be really terrible, right? | 13:39 |
| dansmith | performance-wise | 13:39 |
| bauzas | dansmith: not, because we would actually set it *before* | 13:40 |
| bauzas | dansmith: I'm very against loading in the scheduler | 13:40 |
| dansmith | okay | 13:40 |
| bauzas | dansmith: but there is https://github.com/openstack/nova/blob/master/nova/objects/request_spec.py#L133-L144 | 13:40 |
| bauzas | dansmith: so, that's fine | 13:40 |
| bauzas | dansmith: because then, we would set the filed | 13:40 |
| bauzas | field | 13:40 |
| *** rushil has quit IRC | 13:41 | |
| bauzas | dansmith: so, no lazy-load occuring | 13:41 |
| bauzas | dansmith: the idea to provide a lazy-load is to make sure that this field is correctly populated if needed | 13:41 |
| dansmith | okay | 13:41 |
| *** ctrath has joined #openstack-nova | 13:46 | |
| *** Sree has quit IRC | 13:46 | |
| *** rlrossit has joined #openstack-nova | 13:47 | |
| *** Sree has joined #openstack-nova | 13:47 | |
| *** doude_ has joined #openstack-nova | 13:48 | |
| *** kiran-r has quit IRC | 13:48 | |
| *** doude has quit IRC | 13:48 | |
| *** doude_ has quit IRC | 13:49 | |
| *** annegentle has quit IRC | 13:49 | |
| *** rdopiera has quit IRC | 13:50 | |
| *** rdopiera has joined #openstack-nova | 13:51 | |
| *** rdopiera has quit IRC | 13:51 | |
| *** Sree has quit IRC | 13:51 | |
| *** rdopiera has joined #openstack-nova | 13:52 | |
| *** mwagner_lap has joined #openstack-nova | 13:52 | |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova: CONF.allow_resize_on_same_host should check only once in controller https://review.openstack.org/186236 | 13:53 |
| *** karimb has quit IRC | 13:54 | |
| *** karimb has joined #openstack-nova | 13:56 | |
| *** shuquan has quit IRC | 13:57 | |
| *** zhangjn has joined #openstack-nova | 13:59 | |
| *** zhangjn has quit IRC | 13:59 | |
| *** mriedem has joined #openstack-nova | 13:59 | |
| *** zhangjn has joined #openstack-nova | 14:00 | |
| markus_z | mtreinish: Hi Matthew, do you have a second? | 14:01 |
| *** IanGovett has joined #openstack-nova | 14:01 | |
| *** garyk has quit IRC | 14:01 | |
| markus_z | mtreinish: It's about skipping tests in tempest because of bugs: https://github.com/openstack/tempest/blob/master/tempest/api/compute/admin/test_servers.py#L148 | 14:02 |
| *** doude has joined #openstack-nova | 14:02 | |
| *** doude has quit IRC | 14:02 | |
| *** doude has joined #openstack-nova | 14:02 | |
| *** rdopiera has quit IRC | 14:03 | |
| *** Kennan has quit IRC | 14:03 | |
| *** Kennan2 has joined #openstack-nova | 14:03 | |
| markus_z | mtreinish: Bug 1240043 is solved and would expect that the skip can be removed after that. | 14:03 |
| openstack | bug 1240043 in OpenStack Compute (nova) "get_server_diagnostics must define a hypervisor-independent API" [High,Fix released] https://launchpad.net/bugs/1240043 - Assigned to Gary Kotton (garyk) | 14:03 |
| *** deepakcs has quit IRC | 14:03 | |
| markus_z | mtreinish: It's not a complaint, I'm just curious how it is supposed to be. | 14:04 |
| *** bnemec has joined #openstack-nova | 14:05 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-novaclient: Add method for better random name https://review.openstack.org/207829 | 14:05 |
| dansmith | alaski: your patch to fix the cells test failed the cells test | 14:06 |
| dansmith | alaski: but it's failing in cleanup: Details: {u'code': 400, u'message': u'Invalid volume: Volume still has 1 dependent snapshots.'} | 14:07 |
| *** burgerk has joined #openstack-nova | 14:07 | |
| alaski | hmm, not sure exactly what that means | 14:07 |
| *** rdopiera has joined #openstack-nova | 14:07 | |
| dansmith | alaski: https://jenkins04.openstack.org/job/gate-tempest-dsvm-cells/975/console | 14:08 |
| *** afazekas__ has quit IRC | 14:08 | |
| dansmith | alaski: trying to delete a volume that has a snapshot | 14:08 |
| *** zhangjn has quit IRC | 14:09 | |
| *** jroll has quit IRC | 14:09 | |
| *** jroll has joined #openstack-nova | 14:09 | |
| *** markvoelker has quit IRC | 14:09 | |
| *** lyanchih has joined #openstack-nova | 14:10 | |
| dims | dansmith: rlrossit: looks like we should not recheck until g-r lands for netaddr (https://review.openstack.org/#/c/218720/) and nova's requirements gets updated. should we pre-emptively file a review for nova requirements and not wait for the bot proposed review? | 14:10 |
| openstackgerrit | Noel Nelson Dsouza proposed openstack/nova: Delete flavor should raise error, if it is inuse https://review.openstack.org/217674 | 14:10 |
| dims | sdague: ^^ | 14:10 |
| *** eharney has quit IRC | 14:11 | |
| dansmith | dims: I'm not in a huge rush, as I think that the things that need this will continue to land across FF, since they're RPC bump type maintenance tasks | 14:11 |
| mtreinish | markus_z: push a patch to remove the skip | 14:11 |
| *** boris-42 has joined #openstack-nova | 14:11 | |
| mtreinish | markus_z: the unskips sometimes fall through the cracks | 14:11 |
| dims | dansmith: ack. let nature take its course then :) | 14:11 |
| mtreinish | markus_z: there is a utility skip_tracker which is packaged with tempest-lib that'll detect skips for closed bugs | 14:11 |
| dansmith | dims: yeah, at this point I think we're okay | 14:12 |
| sdague | dims: I don't know that nova is impacted by the netaddr issue | 14:12 |
| sdague | at least, I haven't found an instance of it yet | 14:12 |
| mtreinish | markus_z: but I guess no one bothered to run it in a while :( | 14:12 |
| sdague | neutron, heat, and horizon unit tests are/were | 14:12 |
| *** rushil_ has quit IRC | 14:12 | |
| markus_z | mtreinish: Cool, I'll look for the tool. Something else that I've learned. | 14:13 |
| sdague | yeh, so this popped up in QA as well, but the cells test got to > 50% in *gate* last week - http://graphite.openstack.org/render/?from=-200hours&height=500&until=now&width=800&bgcolor=ffffff&fgcolor=000000&yMax=100&yMin=0&target=color(alias(movingAverage(asPercent(stats.zuul.pipeline.gate.job.gate-tempest-dsvm-cells.FAILURE,sum(stats.zuul.pipeline.gate.job.gate-tempest-dsvm-cells.{SUCCESS,FAILURE})),%2712hours%27),%20%27gate-tempest-dsvm-cells% | 14:13 |
| sdague | 27),%27orange%27) | 14:13 |
| dims | sdague: tempest-dsvm-full tests seem to hit it too - http://logs.openstack.org/92/139992/36/check/gate-tempest-dsvm-full/d0c59d0/console.html#_2015-08-31_08_02_42_671 | 14:13 |
| sdague | which was part of the carnage | 14:13 |
| openstackgerrit | Alex Xu proposed openstack/nova: Add warning log when deprecated v2 and v3 code get used https://review.openstack.org/218103 | 14:13 |
| markus_z | mtreinish: I found (manually) some other obsolete skips too. Would you recommend to make one patch for all findings or one patch per finding? | 14:13 |
| sdague | dims: ok, so "invalid mac address" is addressed | 14:14 |
| alex_xu | sdague: ^ | 14:14 |
| sdague | that was netaddr 0.7.16 and they fixed the upstream bug so we're on 0.7.17 | 14:14 |
| sdague | there are other interfaces changes with really small (/31 /32) net definitinos | 14:14 |
| mtreinish | markus_z: so I did that in the past, and I've found that a few either bit rotted or the bugs weren't fixed | 14:14 |
| dims | sdague: i see. thanks. just making sure i did not miss anything | 14:15 |
| mtreinish | markus_z: so I'd probably do 1 patch to unskip them all | 14:15 |
| mtreinish | markus_z: and then prune from there | 14:15 |
| sdague | which are hitting unit tests on neutron / horizon | 14:15 |
| alex_xu | sdague: and this patch https://review.openstack.org/214601, then those are last two patches for api, hope you can take a look at when you have time, thanks! | 14:15 |
| *** yuntongjin has quit IRC | 14:16 | |
| markus_z | mtreinish: Alright, I'll do that. | 14:16 |
| *** zhangjn has joined #openstack-nova | 14:17 | |
| *** mpaolino has joined #openstack-nova | 14:17 | |
| jaypipes | ndipanov: rebase needed on https://review.openstack.org/#/c/218385/, FYI. | 14:18 |
| ndipanov | jaypipes, yeah I know but working on the last patch in the serios | 14:19 |
| *** annegentle has joined #openstack-nova | 14:19 | |
| ndipanov | and waiting for the lower ones to get through the gate | 14:20 |
| jaypipes | ndipanov: no worries, thanks man! | 14:20 |
| ndipanov | how does it look to you btw jaypipes ? | 14:20 |
| jaypipes | ndipanov: good. I +W'd the first two in the series. | 14:20 |
| *** ildikov has quit IRC | 14:20 | |
| ndipanov | awesomeness | 14:20 |
| ndipanov | jaypipes, I especially wanted your eyes on https://review.openstack.org/#/c/218500/2/nova/compute/manager.py | 14:21 |
| *** zhangjn has quit IRC | 14:21 | |
| *** gholler has joined #openstack-nova | 14:22 | |
| *** tonytan4ever has joined #openstack-nova | 14:22 | |
| *** r-daneel has joined #openstack-nova | 14:22 | |
| *** tpeoples has joined #openstack-nova | 14:23 | |
| *** gholler has quit IRC | 14:24 | |
| *** gholler has joined #openstack-nova | 14:24 | |
| mriedem | rlrossit: https://bugs.launchpad.net/tempest/+bug/1490590 | 14:25 |
| openstack | Launchpad bug 1490590 in tempest "Tempest needs a hacking check that blocks testtools.skip usage" [Low,Confirmed] | 14:25 |
| mriedem | since you're not in the -qa channel | 14:25 |
| rlrossit | ohboyohboy more hacking checks | 14:25 |
| rlrossit | I'm on it boss | 14:26 |
| mtreinish | markus_z: just fyi: http://paste.openstack.org/show/434951/ | 14:26 |
| mtreinish | although it looks like the logic for multiple affecting projects is a bit off | 14:26 |
| mtreinish | that last bug is still open on nova (closed on tempest) | 14:26 |
| *** gholler has quit IRC | 14:26 | |
| *** electrocucaracha has joined #openstack-nova | 14:26 | |
| *** gholler has joined #openstack-nova | 14:27 | |
| *** vivekd has quit IRC | 14:27 | |
| markus_z | mtreinish: Thanks, I'll check them manually in the issue tracker to make it sure. | 14:27 |
| openstackgerrit | Alexander Schmidt proposed openstack/nova: libvirt: Acquire TCP ports for console during live migration https://review.openstack.org/215102 | 14:28 |
| mriedem | rlrossit: although maybe jswarren should take that one | 14:28 |
| *** baoli_ has quit IRC | 14:29 | |
| *** gholler has quit IRC | 14:29 | |
| *** baoli has joined #openstack-nova | 14:30 | |
| rlrossit | mriedem: I can hold off. Just let me know if I should get to work on it | 14:30 |
| mriedem | nah, let john do it, i sent it to him | 14:30 |
| *** lpetrut has quit IRC | 14:32 | |
| *** claudiub has quit IRC | 14:32 | |
| *** hparekh2 has quit IRC | 14:34 | |
| *** Nikolay_St has quit IRC | 14:34 | |
| jaypipes | ndipanov: done | 14:34 |
| *** ildikov has joined #openstack-nova | 14:36 | |
| dansmith | ndipanov: I'm throwing some comments on that set | 14:38 |
| dansmith | not -1ing because they're already approved | 14:38 |
| ndipanov | dansmith, should they be -1ed? | 14:38 |
| dansmith | I just kinda wish some of the things were different, nothing technically wrong :) | 14:39 |
| ndipanov | ok let me see the comments first | 14:39 |
| *** mriedem is now known as mriedem_away | 14:39 | |
| dansmith | there might be reasons why, I haven't made it to the third one yet | 14:40 |
| dansmith | basically, it seems like adding create/destroy/etc on the context object isn't really useful and we should just always load/save it through the instance | 14:40 |
| *** andreykurilin__ has joined #openstack-nova | 14:41 | |
| dansmith | https://review.openstack.org/#/c/218385/2/nova/compute/claims.py,cm | 14:41 |
| *** apuimedo has joined #openstack-nova | 14:42 | |
| dansmith | that creates the migration by itself, but it seems awkward to do that when we've got the actual instance in the caller | 14:42 |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova: Add UniqueConstraint for quota usages table https://review.openstack.org/212908 | 14:42 |
| dansmith | after you do that create, | 14:42 |
| dansmith | your instance doesn't have .migration_context | 14:42 |
| dansmith | so later if you passed it along to someone else, they might expect it to be there | 14:42 |
| dansmith | and if you set it on the instance separately, then the next .save() on the instance will have it be dirty | 14:42 |
| ndipanov | dansmith, yes I was thinking about the same thing | 14:43 |
| *** lpetrut has joined #openstack-nova | 14:43 | |
| ndipanov | but if it ever gets accessed on the instance object it will get loaded | 14:43 |
| ndipanov | so didn't seem lika a big deal | 14:43 |
| dansmith | it won't because it will be =None because when the instance was fetched, it had no context | 14:44 |
| ndipanov | dansmith, orly | 14:44 |
| dansmith | so if you pass the instance on from this point, instance.migration_context will be None and won't re-load | 14:44 |
| ndipanov | that's a problem good poing | 14:44 |
| ndipanov | well not for the current code | 14:44 |
| ndipanov | hmmm | 14:44 |
| dansmith | I just think it's a lot more consistent if we only ever set it on the instance, and handle the create of it by doing a .save() on the instance, | 14:45 |
| dansmith | and delete it by setting =None and .save() again | 14:45 |
| ndipanov | the reason why I probably mistakenly didn't want to do that | 14:45 |
| sdague | dansmith: so, I looked into the DHCPNAK issue that clarkb thought he found in the grenade jobs. It's primarily unrelated. The DHCPNAK is actually correctly NAKing random dhcp requests from unrelated test nodes in that region | 14:45 |
| ndipanov | is that I didn't want to make it look like it's the same thing as other stuff on the instance | 14:45 |
| ndipanov | it's kind of an internal implementation detail of how migrations work | 14:46 |
| dansmith | ndipanov: the numa and pci_requests objects are confusing in this regard, I think, so I would just like to not add more confusing stuff | 14:46 |
| ndipanov | dansmith, yar | 14:46 |
| dansmith | sdague: ... like from other tenants? | 14:46 |
| sdague | well, from other guests | 14:47 |
| sdague | in our tenant | 14:47 |
| sdague | other test jobs | 14:47 |
| dansmith | sdague: ah, okay | 14:47 |
| *** eharney has joined #openstack-nova | 14:48 | |
| *** ftersin has joined #openstack-nova | 14:48 | |
| openstackgerrit | Chuck Carmack proposed openstack/nova: Make the server group invalid format message more verbose https://review.openstack.org/218468 | 14:48 |
| *** achanda has joined #openstack-nova | 14:49 | |
| *** zhangjn has joined #openstack-nova | 14:49 | |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova: Add UniqueConstraint for quota usages table https://review.openstack.org/212908 | 14:50 |
| *** moshele has quit IRC | 14:51 | |
| *** dims has quit IRC | 14:51 | |
| ndipanov | dansmith, so that is only the instance object changing patch that needs to be changed right? | 14:52 |
| ndipanov | -1 or 2 that oen and I will re-upload it soon | 14:52 |
| dansmith | ndipanov: you might as well remove the remotable create, etc methods from the object too right? | 14:52 |
| *** markvoelker has joined #openstack-nova | 14:53 | |
| ndipanov | dansmith, well I could but that's pedantic at that point no? | 14:53 |
| dansmith | ndipanov: if they're not used? | 14:54 |
| ndipanov | ah right... | 14:54 |
| lyanchih | Hi, recently I report a bug https://bugs.launchpad.net/nova/+bug/1489744 . It is talking about swap volume will stay in incorrect status. When I try to fix, I found it is cause by nova don’t send final attach and detach action to cinder. | 14:54 |
| openstack | Launchpad bug 1489744 in OpenStack Compute (nova) "Swapping volume will not result in correct status" [Undecided,New] - Assigned to Chung Chih, Hung (lyanchih) | 14:54 |
| openstackgerrit | Dan Smith proposed openstack/nova: objects: Add MigrationContext object https://review.openstack.org/216732 | 14:54 |
| *** markvoelker has quit IRC | 14:54 | |
| dansmith | ndipanov: I just did a commit message change to pop it out immediately.. ping me when you have it up and I'll review right away | 14:55 |
| *** markvoelker has joined #openstack-nova | 14:55 | |
| lyanchih | Nova though these actions should been done in cinder. But I don’t find any relation code at cinder. | 14:55 |
| dansmith | ndipanov: also, I think that as soon as mitaka opens, we can stop doing the relationship mapping crap | 14:56 |
| *** ssurana has quit IRC | 14:57 | |
| dansmith | ndipanov: as lemming will do all requests by manifest | 14:57 |
| *** jistr is now known as jistr|call | 14:57 | |
| dansmith | I'm pretty (damn) excited for that day | 14:57 |
| *** cristicalin has quit IRC | 14:58 | |
| lyanchih | Therefore I want to ask for help. What should I do in this situation? Should I ask cinder member in cinder irc? | 14:59 |
| *** sudipto has joined #openstack-nova | 15:01 | |
| *** baoli has quit IRC | 15:01 | |
| *** baoli has joined #openstack-nova | 15:02 | |
| *** s_amann has quit IRC | 15:02 | |
| *** dims has joined #openstack-nova | 15:03 | |
| *** s_amann has joined #openstack-nova | 15:03 | |
| jaypipes | lyanchih: yes. please talk with thingee (Mike Perez) in #openstack-cinder. They are aware of these issues and he can point you to some patches that try to address. | 15:03 |
| ndipanov | dansmith, ok thanks | 15:04 |
| *** hemnafk has quit IRC | 15:04 | |
| *** bnemec has quit IRC | 15:04 | |
| *** achanda has quit IRC | 15:05 | |
| lyanchih | jaypipes: Ok, thank you. I will ask at there. | 15:05 |
| *** shoutm has quit IRC | 15:05 | |
| *** Murali has quit IRC | 15:05 | |
| *** bnemec has joined #openstack-nova | 15:09 | |
| openstackgerrit | Brian Elliott proposed openstack/nova: Add migration policy to upgrades devref https://review.openstack.org/218362 | 15:09 |
| *** tjones has joined #openstack-nova | 15:11 | |
| *** dave-mccowan has quit IRC | 15:11 | |
| openstackgerrit | Balazs Gibizer proposed openstack/nova: Add functional test for server group https://review.openstack.org/147514 | 15:13 |
| lyanchih | Hi, please help review our patch https://review.openstack.org/#/c/208268/ | 15:13 |
| *** apuimedo is now known as apuimedo|away | 15:13 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Enable options for oslo.reports https://review.openstack.org/216917 | 15:13 |
| *** tjones has left #openstack-nova | 15:14 | |
| *** electrocucaracha has quit IRC | 15:17 | |
| *** electrocucaracha has joined #openstack-nova | 15:18 | |
| *** prometheanfire has joined #openstack-nova | 15:19 | |
| prometheanfire | so, I know force_config_drive is going away in liberty, but what happens to what it was doing? is it always forced on now or what? | 15:21 |
| bauzas | jaypipes: I'm in a call atm, but I would appreciate if you could just reply again to my comments in https://review.openstack.org/#/c/213591/9 ? | 15:22 |
| bauzas | jaypipes: I tried to be more verbose to explain my questions | 15:22 |
| openstackgerrit | Simona Iuliana Toader proposed openstack/nova: Hyper-V vTPM Devices https://review.openstack.org/218642 | 15:23 |
| *** thedodd has joined #openstack-nova | 15:23 | |
| *** claudiub has joined #openstack-nova | 15:23 | |
| *** dave-mccowan has joined #openstack-nova | 15:25 | |
| *** VikasC has joined #openstack-nova | 15:25 | |
| *** jaypipes has quit IRC | 15:26 | |
| *** haomaiwa_ has quit IRC | 15:29 | |
| *** david-ly_ has joined #openstack-nova | 15:31 | |
| *** sudipto_ has joined #openstack-nova | 15:32 | |
| *** david-ly_ is now known as david-lyle_ | 15:33 | |
| *** sudipto has quit IRC | 15:33 | |
| *** david-lyle has quit IRC | 15:33 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/216968 | 15:35 |
| *** david-lyle_ is now known as david-lyle | 15:35 | |
| markus_z | prometheanfire: AFAIK only the value "always" is deprecated for "force_config_drive". | 15:36 |
| markus_z | prometheanfire: The default value will be "True" instead. | 15:37 |
| markus_z | prometheanfire: What makes you say that this option is going away? | 15:37 |
| *** lyanchih has quit IRC | 15:38 | |
| *** alex_klimov has quit IRC | 15:38 | |
| *** Nikolay_St has joined #openstack-nova | 15:39 | |
| *** IanGovett1 has joined #openstack-nova | 15:43 | |
| *** IanGovett has quit IRC | 15:43 | |
| *** Piet has joined #openstack-nova | 15:44 | |
| prometheanfire | markus_z: the docs | 15:45 |
| prometheanfire | markus_z: http://docs.openstack.org/kilo/config-reference/content/list-of-compute-config-options.html | 15:45 |
| prometheanfire | markus_z: does true mean that it's always attaching the config drive then? | 15:46 |
| *** sudipto_ has quit IRC | 15:46 | |
| markus_z | prometheanfire: Yes, "True" means exactly that. | 15:47 |
| markus_z | prometheanfire: https://github.com/openstack/nova/blob/master/nova/virt/configdrive.py#L179 | 15:47 |
| prometheanfire | ah, cool | 15:47 |
| prometheanfire | so always basically means true? | 15:47 |
| prometheanfire | or was always meaning user couldn't disdable it with 'nova boot config_drive=false' or whatever? | 15:48 |
| *** rushil has joined #openstack-nova | 15:48 | |
| markus_z | prometheanfire: IIUC, "always" == "True". | 15:49 |
| prometheanfire | right | 15:49 |
| prometheanfire | cool, thanks | 15:49 |
| markus_z | prometheanfire: You're welcome | 15:49 |
| *** amotoki has quit IRC | 15:50 | |
| *** VikasC has quit IRC | 15:50 | |
| *** Sukhdev has joined #openstack-nova | 15:51 | |
| markus_z | prometheanfire: For the sake of completeness, "nova boot config-drive=False" doesn't do anything. Only "True" has an effect and only if the "force_config_drive" option was "False". | 15:52 |
| prometheanfire | ok | 15:53 |
| *** jistr|call is now known as jistr | 15:53 | |
| prometheanfire | nice that we are always presenting the data then | 15:53 |
| *** pbrooko has joined #openstack-nova | 15:53 | |
| prometheanfire | iirc there was a problem if the data changed, the iso wouldn't get regened or something (network updates) | 15:53 |
| *** zhangjn has quit IRC | 15:54 | |
| *** yassine__ has quit IRC | 15:54 | |
| markus_z | You mean an updated of the data on the config drive? | 15:55 |
| prometheanfire | yes | 15:55 |
| markus_z | AFAIK the content should be pretty static, otherwise it doesn't do its one job, to provide data for configuration at boot time. | 15:56 |
| markus_z | When does an update of the data happen? | 15:57 |
| prometheanfire | it should happen if you add/change networking for instance | 15:58 |
| *** diegows has joined #openstack-nova | 15:58 | |
| prometheanfire | attach/detach networks | 15:58 |
| prometheanfire | maybe drives too | 15:58 |
| prometheanfire | cloud-init would have to be re-run though | 15:58 |
| prometheanfire | so not sure how useful really | 15:58 |
| *** annegentle has quit IRC | 15:59 | |
| markus_z | This happens after the boot of the instance is done, which isn't within the scope of the config-drive I guess. But I'm not really sure. | 15:59 |
| prometheanfire | that's running config changes, generally outside of cloud-init's view | 16:00 |
| *** markvoelker_ has joined #openstack-nova | 16:00 | |
| openstackgerrit | jichenjc proposed openstack/nova: API: remove unused parameter https://review.openstack.org/218932 | 16:00 |
| prometheanfire | glad we agree then :D | 16:00 |
| *** MarcF has quit IRC | 16:00 | |
| markus_z | :) | 16:00 |
| *** markvoel_ has joined #openstack-nova | 16:01 | |
| prometheanfire | guess all that would be config management of the guest to be done by either auto-dhcp for addresses or something like ansible/puppet/etc | 16:01 |
| markus_z | prometheanfire: Right, I think so to. These tools are made for this and should do a better job. Config-drive can be a leverage for using that. | 16:03 |
| *** Apoorva has joined #openstack-nova | 16:03 | |
| *** markvoelker has quit IRC | 16:03 | |
| *** dannywilson has joined #openstack-nova | 16:04 | |
| *** baoli has quit IRC | 16:04 | |
| *** sayali has quit IRC | 16:04 | |
| *** jichen has quit IRC | 16:04 | |
| *** IanGovett1 has quit IRC | 16:05 | |
| *** baoli has joined #openstack-nova | 16:05 | |
| prometheanfire | right | 16:05 |
| *** markvoelker_ has quit IRC | 16:05 | |
| *** apuimedo|away has quit IRC | 16:05 | |
| *** baoli has quit IRC | 16:05 | |
| prometheanfire | also, that would mean cloud-init and config management could fight | 16:05 |
| *** IanGovett has joined #openstack-nova | 16:06 | |
| markus_z | And we don't want fights ;) | 16:06 |
| * markus_z has to leave for today. see you | 16:06 | |
| *** sayali has joined #openstack-nova | 16:06 | |
| *** markus_z has quit IRC | 16:06 | |
| prometheanfire | cya | 16:06 |
| *** mriedem_away is now known as mriedem | 16:08 | |
| *** matrohon has quit IRC | 16:08 | |
| *** apuimedo|away has joined #openstack-nova | 16:09 | |
| *** flashgordon is now known as jogo | 16:09 | |
| *** IanGovett1 has joined #openstack-nova | 16:10 | |
| *** IanGovett has quit IRC | 16:10 | |
| *** MarcF has joined #openstack-nova | 16:12 | |
| *** kiran-r has joined #openstack-nova | 16:13 | |
| *** jistr has quit IRC | 16:13 | |
| *** ssurana has joined #openstack-nova | 16:14 | |
| *** armax has joined #openstack-nova | 16:15 | |
| *** ssurana has left #openstack-nova | 16:15 | |
| *** mpaolino has quit IRC | 16:16 | |
| *** vishwanathj has joined #openstack-nova | 16:18 | |
| *** vivekd has joined #openstack-nova | 16:18 | |
| *** jaypipes has joined #openstack-nova | 16:22 | |
| *** skylerberg has joined #openstack-nova | 16:22 | |
| *** salv-orl_ has joined #openstack-nova | 16:25 | |
| *** unicell has quit IRC | 16:25 | |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: RT: Migration resource tracking uses migration context https://review.openstack.org/218938 | 16:25 |
| *** salv-orlando has quit IRC | 16:28 | |
| *** mdrabe has quit IRC | 16:28 | |
| *** markvoel_ has quit IRC | 16:31 | |
| *** ociuhandu has joined #openstack-nova | 16:31 | |
| *** markvoelker has joined #openstack-nova | 16:31 | |
| *** markvoelker_ has joined #openstack-nova | 16:32 | |
| *** lifeless has quit IRC | 16:34 | |
| *** annegentle has joined #openstack-nova | 16:35 | |
| *** karimb has quit IRC | 16:36 | |
| *** markvoelker has quit IRC | 16:36 | |
| edleafe | johnthetubaguy: around? Question about v3 deprecation | 16:37 |
| *** markvoelker_ has quit IRC | 16:37 | |
| *** gszasz has quit IRC | 16:38 | |
| *** sahid has quit IRC | 16:41 | |
| *** lifeless has joined #openstack-nova | 16:41 | |
| *** mdrabe has joined #openstack-nova | 16:41 | |
| *** s_amann has quit IRC | 16:44 | |
| *** meghal has joined #openstack-nova | 16:44 | |
| *** nic has joined #openstack-nova | 16:46 | |
| *** davideagnello has joined #openstack-nova | 16:47 | |
| *** davideagnello has quit IRC | 16:49 | |
| *** romainh has left #openstack-nova | 16:50 | |
| *** yamahata has joined #openstack-nova | 16:50 | |
| *** davideagnello has joined #openstack-nova | 16:50 | |
| openstackgerrit | garyk proposed openstack/nova: DB: add support for volume_type by BDMS https://review.openstack.org/139992 | 16:51 |
| openstackgerrit | garyk proposed openstack/nova: Object: add volume_type_id to block_device objects https://review.openstack.org/139994 | 16:52 |
| *** pbrooko has quit IRC | 16:52 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Remove duplicate NullHandler test fixture https://review.openstack.org/218950 | 16:53 |
| *** lpetrut has quit IRC | 16:53 | |
| *** prometheanfire has left #openstack-nova | 16:54 | |
| *** dims has quit IRC | 16:55 | |
| *** salv-orl_ has quit IRC | 16:56 | |
| *** cfriesen has joined #openstack-nova | 16:58 | |
| *** vishwanathj has quit IRC | 17:00 | |
| cfriesen | any chance of getting https://review.openstack.org/#/c/210689/ some attention? It's a small fix to return the "correct" error message for attempted operations on locked instances. | 17:00 |
| *** earlephilhower has joined #openstack-nova | 17:02 | |
| *** dannywilson has quit IRC | 17:02 | |
| *** exploreshaifali has joined #openstack-nova | 17:04 | |
| *** derekh has quit IRC | 17:05 | |
| *** IanGovett1 has quit IRC | 17:08 | |
| *** kiran-r has quit IRC | 17:10 | |
| *** tonytan4ever has quit IRC | 17:12 | |
| *** unicell has joined #openstack-nova | 17:12 | |
| *** rdopiera has quit IRC | 17:15 | |
| *** achanda has joined #openstack-nova | 17:18 | |
| *** nic has quit IRC | 17:18 | |
| *** garyk has joined #openstack-nova | 17:20 | |
| *** aix has quit IRC | 17:21 | |
| mriedem | cfriesen: +W | 17:22 |
| *** vivekd has quit IRC | 17:23 | |
| *** alexschm has quit IRC | 17:23 | |
| *** mjturek1 has quit IRC | 17:23 | |
| *** IanGovett has joined #openstack-nova | 17:24 | |
| *** nic has joined #openstack-nova | 17:24 | |
| mriedem | dansmith: https://review.openstack.org/#/c/218362/ is all yours | 17:25 |
| dansmith | okay, in a sec | 17:25 |
| *** ftersin has left #openstack-nova | 17:25 | |
| garyk | dansmith: did you see the mail thread about https://review.openstack.org/190047 - on the review you wanted discussion and i did not see you chime in. can we unblock this. it is a real issue that we have in production | 17:27 |
| mriedem | zzzeek: in https://review.openstack.org/#/c/212908/ - i'm wondering if you have some sqlalchemy-fu where we could detect things that would fail creating a unique constraint and reduce those before actually doing the UC.create() call | 17:28 |
| dansmith | garyk: okay, the discussion seemed to be happening, but I didn't follow the end of it, I'll look in a bit | 17:28 |
| *** dims has joined #openstack-nova | 17:29 | |
| *** vilobhmm1 has joined #openstack-nova | 17:29 | |
| zzzeek | mriedem: well you’d want to run a SELECT statement that looks for dupes | 17:29 |
| zzzeek | mriedem: is that the part you want to see ? | 17:29 |
| *** dims_ has joined #openstack-nova | 17:30 | |
| *** dims has quit IRC | 17:30 | |
| *** emagana has joined #openstack-nova | 17:30 | |
| *** mjturek1 has joined #openstack-nova | 17:30 | |
| *** ildikov has quit IRC | 17:31 | |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova-specs: Libvirt hardware policy from libosinfo https://review.openstack.org/218962 | 17:31 |
| *** meghal has quit IRC | 17:32 | |
| mriedem | zzzeek: yeah | 17:32 |
| dansmith | garyk: I seem to be missing where that thread came to any sort of conclusion that indicates the patch should be merged | 17:33 |
| *** meghal has joined #openstack-nova | 17:34 | |
| zzzeek | mriedem: so suppose you had UNIQUE on columns “x” and “y”. you can find dupes as SELECT count(*) FROM table GROUP BY x, y HAVING count(*) > 1 | 17:35 |
| *** baoli has joined #openstack-nova | 17:35 | |
| garyk | dansmith: i am not sure what to say. i state a number of cases where it can and does happen. this is with the vmware driver and other drivers. | 17:35 |
| *** sachin has joined #openstack-nova | 17:35 | |
| garyk | the fact tha matt is not convinced and you are against it does not help. the fact that the patch is just logging instead of doing something drastic is an issue. | 17:36 |
| mriedem | zzzeek: is that all possible via sqlalchemy? or do we have to do straight sql execution? | 17:36 |
| *** tonytan4ever has joined #openstack-nova | 17:36 | |
| garyk | and as you said the hypervisor should be the source of truth - and that is what is happening - it is the source of truth, but the compute node decides to shut down an instance | 17:36 |
| garyk | did i not manage to get that point across? | 17:37 |
| zzzeek | mriedem: heh. there is no SELECT we dont do :) | 17:37 |
| *** baoli has quit IRC | 17:37 | |
| mriedem | zzzeek: i can't beat your low low prices? :) | 17:37 |
| zzzeek | mriedem: well here’s the thing, you want to do this in a migration? let’s go to the chorus. “what about online schema migrations?” | 17:37 |
| garyk | dansmith: i think that the same can happen with libvurt, for example if the libvirt process is being restarted or hangs for a few seconds whilt the periodic task is running | 17:37 |
| garyk | that can certainly happen | 17:38 |
| *** baoli has joined #openstack-nova | 17:38 | |
| garyk | libvirt | 17:38 |
| mriedem | garyk: if libvirt restarts the nova driver should punt | 17:38 |
| zzzeek | mriedem: a core SELECT looks like stmt = select([func.count(‘*’)]).group_by(MyModel.x, MyModel.y).having(func.count(‘*’) > 1); if connection.scalar(stmt): boom | 17:38 |
| mriedem | zzzeek: ah, nice | 17:38 |
| *** rlrossit has left #openstack-nova | 17:39 | |
| garyk | mriedem: should punt for what? so nova compute should restart? | 17:39 |
| mriedem | garyk: no, when asked about the power_state of a vm, if the nova libvirt driver can't connect to libvirtd to determine that, it should raise an error | 17:39 |
| garyk | ok, so we have a different us case here. so for hyperv and vmware this may be different. | 17:40 |
| garyk | so do we need to block these drivers - this change is just logging something and not doing anything else - backwards compatible. | 17:40 |
| mriedem | as in the instance moved from vmware black box A to vmware black box B and nova doesn't know about said black boxes? | 17:40 |
| claudiub | lifecycle events? | 17:41 |
| garyk | mriedem: the general use case for VMware is to enable DRS and HA. so instances can be moved from one box to anoterh | 17:41 |
| garyk | that should not have the compute node shutting down the instance | 17:41 |
| claudiub | if lifecycle events are implemented, nova compute manager should notice that instances moved from one node to another | 17:42 |
| mriedem | garyk: can't the nova vmware driver detect that the instance is moving? | 17:42 |
| garyk | this is an edge case when one of the box failed - say for example a power outage. happsn. no? | 17:42 |
| garyk | the HA moves that instance to another box and the service continues | 17:42 |
| garyk | no? | 17:42 |
| mriedem | garyk: no one is arguing that shit doesn't happen | 17:42 |
| garyk | or do we want users to build applicatiosn that can handle random shutdowns | 17:42 |
| garyk | yeah, but lets not have the copmpute node behaving as cattle | 17:42 |
| garyk | that is my point | 17:42 |
| mriedem | garyk: i think the point people are trying to make is the vmware driver should not be returning bogus information to the task | 17:42 |
| *** lpetrut has joined #openstack-nova | 17:43 | |
| garyk | kind of stuff breaks, but lets not have the compute node add more havic to that | 17:43 |
| garyk | mriedem: it is not returning bogus data | 17:43 |
| *** ig0r__ has joined #openstack-nova | 17:43 | |
| garyk | the instance is down at point A when the host is down and it is up and point B when the instance has been reolocated | 17:43 |
| garyk | the fact that the compute node incorrectly decides that it should be down is where the problem lies | 17:43 |
| garyk | do you not understand that? | 17:44 |
| garyk | or am i unable to explain it | 17:44 |
| alaski | garyk: how does the compute node decide that it's down? | 17:44 |
| alaski | the virt driver is telling it that it is right? | 17:44 |
| garyk | alaski: the compute node queries the VC. the VC sees that it down and says it is down | 17:44 |
| garyk | next perioedic iteration is is up | 17:44 |
| *** SourabhP has joined #openstack-nova | 17:45 | |
| cfriesen | garyk: mriedem: I think the point here is that the vmware hypervisor is essentially changing the state autonomously without going through the nova APIs, so presumably we'd want to update the nova state to reflect what the hypervisor is telling us. | 17:45 |
| garyk | as the HA moved it from the hosed host to another healthy one | 17:45 |
| mriedem | garyk: can't the vmware driver determine that the instance is down because the host is also down? | 17:45 |
| garyk | mriedem: no, it cannot | 17:45 |
| mriedem | and raise an error to the task to ignore htat instnace? | 17:45 |
| garyk | it is not aware of the hosts in the cluster | 17:45 |
| cfriesen | mriedem: the instance is actually down at the time of the periodic task scan though...so it's valid to report that it's down | 17:45 |
| mriedem | cfriesen: sure, and then it's back up, but nova will think it's still down and stop it | 17:46 |
| mriedem | i get that | 17:46 |
| garyk | cfriesen: that is what the patch that i added does - it logs that it is down and just does not shut it down when it comes back up | 17:46 |
| *** ig0r_ has quit IRC | 17:46 | |
| mriedem | garyk: your patch actualy doesn't handle that case | 17:46 |
| mriedem | which is why i just pointed out again in the change | 17:46 |
| garyk | mriedem: that is exactly what my patch handles | 17:46 |
| mriedem | no, your patch ignores the case when it's actually down | 17:46 |
| mriedem | i'm saying, | 17:46 |
| mriedem | it's not handling the case where it's actually running but nova turns it off | 17:47 |
| garyk | mriedem: no, when it is down at point A the instance is marked as down and the periodic task continues | 17:47 |
| mriedem | i think it's totally valid for nova to call the stop API when the nova db say it's active but the hypervisor says it's stopped | 17:47 |
| garyk | that is what the code does today | 17:47 |
| *** mjturek1 has quit IRC | 17:47 | |
| garyk | mriedem: but the point A when the instance is down the DB is updated from running to stooped | 17:48 |
| mriedem | garyk: tell me where i'm wrong here: (12:38:46 PM) zzzeek: mriedem: a core SELECT looks like stmt = select([func.count(‘*’)]).group_by(MyModel.x, MyModel.y).having(func.count(‘*’) > 1); if connection.scalar(stmt): boom | 17:48 |
| garyk | then the next iteration when it is up it is stopped | 17:48 |
| mriedem | damn | 17:48 |
| mriedem | here https://review.openstack.org/#/c/190047/12/nova/compute/manager.py | 17:48 |
| *** aix has joined #openstack-nova | 17:48 | |
| garyk | mriedem: lets start again: | 17:49 |
| garyk | 1. periodic task runs at point A. every is fine. | 17:49 |
| garyk | 2. host A goes down | 17:49 |
| cfriesen | mriedem: vm_power_state comes from the hypervisor, no? | 17:49 |
| garyk | 3. periodic task runs at point B. instance X is down. it is marked as stooped in the DB by the periodic task | 17:49 |
| cfriesen | so the hypervisor is reporting it's down, but the DB has it as active | 17:50 |
| garyk | 4. vmware VC detects that host A is down and reoltaes all instances to other hosts in cluster and restarts those instances | 17:50 |
| garyk | 5. periodic tasks runs and hyperviosr says instance is running - this does not match the DB state | 17:50 |
| garyk | so perdioc tasks shuts down the instance | 17:50 |
| mriedem | garyk: yes, i know all this | 17:50 |
| mriedem | #3 is totally valid | 17:50 |
| openstackgerrit | lalit dagre proposed openstack/nova: Flavor object which support both DB with active migration https://review.openstack.org/213041 | 17:50 |
| mriedem | what i'm saying is, | 17:50 |
| mriedem | if you were going to ignore something, i'd ignore #5 | 17:51 |
| mriedem | since #5 is the one that ignores the hypervisor for the nova db | 17:51 |
| mriedem | however, | 17:51 |
| alaski | garyk: my question on this is still if host A is down why it returns info to the periodic task to say the instance is stopped | 17:51 |
| mriedem | then nova says the thing is not running when it is | 17:51 |
| *** Sukhdev has quit IRC | 17:51 | |
| garyk | alaski: at that point in time the instance is actually down. | 17:51 |
| mriedem | i'd think #5 would actually do like a reboot or something | 17:51 |
| openstackgerrit | Joshua Harlow proposed openstack/nova: libvirt: Use psutil and TERM to KILL backoff loop https://review.openstack.org/218597 | 17:52 |
| garyk | the VC driver returns that the instance is down | 17:52 |
| cfriesen | mriedem: I think where your comment is in the review is actually representing step 3 | 17:52 |
| alaski | garyk: butif the host is down, return an error, not say that the instance is down | 17:52 |
| garyk | mriedem: i do not see what it would do a reboot. the hypervisor returns that the instance is running - so now you want to reboot it? sorry i do not understand | 17:52 |
| cfriesen | alaski: but if the host is down, at that point the instance is also down | 17:52 |
| mriedem | cfriesen: no, #3 is L5951 | 17:52 |
| alaski | cfriesen: yes, but if the host is down just indicate that it's not valid to sync power states now | 17:53 |
| garyk | alaski: so you are suggesting that the driver return that the instance is in an error state? | 17:53 |
| mriedem | garyk: reboot it to re-sync the nova db and the hypervisor | 17:53 |
| cfriesen | mriedem: crap, you're right. missed the skipped lines | 17:53 |
| mriedem | alaski: +1 | 17:53 |
| *** mjturek1 has joined #openstack-nova | 17:53 | |
| mriedem | basically, the virt driver should somehow detect sh is f'ed and says skip me for now | 17:53 |
| mriedem | i'm sitting this one out | 17:53 |
| alaski | garyk: not exactly, the driver could return and say "it is not valid to sync power states right now because a host is down and I don't know its power state" | 17:53 |
| cfriesen | alaski: but host down means we *know* the instance is donw | 17:53 |
| mriedem | alaski: s is f'ed | 17:54 |
| mriedem | is the technical term | 17:54 |
| cfriesen | at least until we restart it somewhere | 17:54 |
| garyk | alaski: i will look at that option | 17:54 |
| alaski | sure, except HA means we don't actually know that apparently | 17:54 |
| garyk | alaski: the HA knows that the instance should be moved. | 17:54 |
| alaski | right, meaning the state isn't really known | 17:54 |
| garyk | and it correctly does that. the fact thatthe compute node takes drastic action is not an issue that concerns you guys? | 17:54 |
| mriedem | cfriesen: there is a difference between knowing the instance is down and knowing that it's down but will be moved and up again on another host | 17:54 |
| cfriesen | mriedem: okay, but until it's moved and up again, it's down. :) | 17:55 |
| mriedem | garyk: sure it does, that's why i pushed up the patch to disable lifecycle events | 17:55 |
| cfriesen | mriedem: and that move/restart could fail due to resource shortage | 17:55 |
| garyk | mriedem: but that has even more problems - and you are unable to delete instances | 17:55 |
| cfriesen | mriedem: in which case we'd want it to show as "down", no? | 17:55 |
| mriedem | garyk: why am i unable to delete instances? | 17:55 |
| tojuvone | cfriesen: Would you have nice comment on this spec that continues the mark-host-down to be visible on instance: https://review.openstack.org/#/c/192246/7 | 17:56 |
| garyk | i am not sure if you have encountered that - then the tenants quotas are all used up due to a dodgy compute node logic | 17:56 |
| *** rlrossit has joined #openstack-nova | 17:56 | |
| mriedem | garyk: huh | 17:56 |
| cfriesen | tojuvone: will try. :) | 17:56 |
| mriedem | garyk: if i delete the instance the quota is freed up | 17:56 |
| mriedem | i should always be able to delete an instance | 17:56 |
| garyk | mriedem: i have hit cases when the states are not in sync the api is unable to delete the instances | 17:57 |
| mriedem | garyk: yeah, those are bugs | 17:57 |
| garyk | mriedem: yeah. | 17:57 |
| garyk | mriedem: and so is this | 17:57 |
| garyk | i will look at what alaski suggested. i am still not convinced that it is the right direction | 17:57 |
| garyk | the tenant will see that the instance is in ERROR state and it is actually running. that will also cause issues | 17:58 |
| mriedem | garyk: what alaski said is basically the same as what dansmith said when he -2'd the change | 17:58 |
| garyk | if the host fails then it is really down and if the host restarts then it will really be running | 17:58 |
| cfriesen | mriedem: alaski: if we take the hypervisor to be correct when it reports the VM as down, shouldn't we take it to be correct when it reports the VM to be up? | 17:58 |
| alaski | garyk: don't put the instance in error, just skip the sync and try again when the host is up or the instances are all moved | 17:58 |
| mriedem | cfriesen: that's what i was pointing out in the comments | 17:58 |
| garyk | cfriesen: that is my argument, why is the compute node overriding the hy[pervisor | 17:58 |
| mriedem | alaski: basically treat it as the task_state is not None but in the virt driver | 17:59 |
| cfriesen | mriedem: right | 17:59 |
| *** tojuvone_ has joined #openstack-nova | 17:59 | |
| alaski | mriedem: exactly | 17:59 |
| cfriesen | garyk: see mriedem's comments in the review... | 17:59 |
| mriedem | if the virt driver can detect or assume there is a task being performed due to HA, | 17:59 |
| mriedem | skip it | 17:59 |
| mriedem | like the periodi task does if there is task happening at the top | 17:59 |
| tojuvone_ | cfiresen: thanks :) | 17:59 |
| garyk | ok, i'll look into that | 17:59 |
| tojuvone_ | Also if some other API guys could comment, would be nice :) | 18:00 |
| *** VW_ has quit IRC | 18:00 | |
| mriedem | if nothing else, i'd consider a change that allows an operator to either stop or reboot the instance in the case that the nova db says it's down but the hypervisor says its up | 18:00 |
| *** tojuvone has quit IRC | 18:00 | |
| *** VW_ has joined #openstack-nova | 18:00 | |
| mriedem | like what resume_guests_state_on_host_boot is used for | 18:00 |
| garyk | mriedem: ok | 18:01 |
| *** lucasagomes is now known as lucas-dinner | 18:01 | |
| garyk | mriedem: so for this inad of shutting down we could just use this vr and continue? | 18:02 |
| mriedem | vr? | 18:02 |
| garyk | mriedem: variable | 18:02 |
| mriedem | resume_guests_state_on_host_boot ? | 18:03 |
| mriedem | that config option is a bit confusing in this context | 18:03 |
| mriedem | but i think it's used in the same manner on init | 18:03 |
| *** sachin has quit IRC | 18:04 | |
| *** salv-orlando has joined #openstack-nova | 18:04 | |
| garyk | mriedem: say for example https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L5975 the accoring to the flag dont do the shutdown | 18:05 |
| *** sachin has joined #openstack-nova | 18:05 | |
| *** davideagnello has quit IRC | 18:06 | |
| garyk | mriedem: i'll look at this tomorrow | 18:07 |
| mriedem | garyk: no | 18:07 |
| mriedem | i think in that case the hypervisor says it's down so call the stop api | 18:07 |
| garyk | ok, i wont look at it tomorrow :) | 18:07 |
| mriedem | that makes sense | 18:07 |
| mriedem | garyk: i'm saying this case https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L6027 | 18:08 |
| mriedem | which is what i commented on in the change | 18:08 |
| openstackgerrit | Ed Leafe proposed openstack/nova: Remove 'v3' from extension code https://review.openstack.org/214311 | 18:09 |
| openstackgerrit | Ed Leafe proposed openstack/nova: Remove more 'v3' references from the code https://review.openstack.org/214290 | 18:09 |
| *** salv-orlando has quit IRC | 18:09 | |
| garyk | mriedem: that kind of was addressed by my change - in line 5924 | 18:09 |
| garyk | but not in the way that you wanted me to deal with it | 18:10 |
| mriedem | garyk: that's the wrong place | 18:10 |
| mriedem | garyk: i think it fixes your scenario but in the wrong place | 18:10 |
| mriedem | sec, i'll push up a WIP in a sec here | 18:10 |
| garyk | mriedem: no need, i'll do that | 18:11 |
| openstackgerrit | Chuck Carmack proposed openstack/nova: Fix the usage output of the nova-idmapshift command https://review.openstack.org/218973 | 18:12 |
| *** sachin has quit IRC | 18:13 | |
| *** takedakn has quit IRC | 18:13 | |
| *** sachin has joined #openstack-nova | 18:13 | |
| openstackgerrit | Andrew Laski proposed openstack/nova: Add persistence to the RequestSpec object https://review.openstack.org/211753 | 18:14 |
| *** sachin has quit IRC | 18:16 | |
| *** sachin has joined #openstack-nova | 18:17 | |
| *** sachin has quit IRC | 18:17 | |
| *** sachin has joined #openstack-nova | 18:17 | |
| *** markvoelker has joined #openstack-nova | 18:17 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Allow admin to reboot instances when syncing power states https://review.openstack.org/218975 | 18:19 |
| mriedem | garyk: dansmith: alaski: cfriesen: ^ alternative solution, needs tests | 18:19 |
| *** ildikov has joined #openstack-nova | 18:20 | |
| *** markvoelker has quit IRC | 18:21 | |
| *** sachin has quit IRC | 18:21 | |
| *** sachin has joined #openstack-nova | 18:21 | |
| *** pv_ has joined #openstack-nova | 18:23 | |
| pv_ | hey there, quick question.. whenever I try to boot a nova instance its getting stuck in a BUILD state.. my n-cpu logs show an error in nova.compute.manager: Error: \'versioned_object.namespace\'\n' | 18:24 |
| pv_ | and the same error appears in my n-cond.log in rpc.dispatcher | 18:24 |
| dansmith | pv_: there's more to the error | 18:24 |
| pv_ | ERROR oslo_messaging.rpc.dispatcher [req-f92a3b5a-23db-45d8-aaa3-fe503583bdec None None] Exception during message handling: 'versioned_object.namespace' | 18:25 |
| dansmith | should be a lot above that line | 18:25 |
| pv_ | is the start of a long line of error messages im getting in my conductor logs | 18:25 |
| *** davideagnello has joined #openstack-nova | 18:25 | |
| *** rushil has quit IRC | 18:25 | |
| pv_ | oh yeah there is | 18:25 |
| dansmith | yeah, pastebin please | 18:25 |
| pv_ | kk one sec | 18:25 |
| bauzas | pv_: dansmith: I'm just having the same problem | 18:27 |
| pv_ | http://pastebin.com/uNWvcv26 error message in my n-cpu log | 18:27 |
| pv_ | ill get you the cond log in a sec | 18:28 |
| bauzas | when running local tests | 18:28 |
| bauzas | okay, unrelated to me | 18:28 |
| *** exploreshaifali has quit IRC | 18:28 | |
| bauzas | nevermind | 18:28 |
| dansmith | pv_: yeah, the conductor log will be more helpful | 18:28 |
| dansmith | pv_: are you running master? | 18:28 |
| pv_ | master? haha im not sure what that is | 18:29 |
| pv_ | sorry | 18:29 |
| *** takedakn has joined #openstack-nova | 18:30 | |
| bauzas | dansmith: okay, I think I found my own problem related to that, probably due to an old o.voo | 18:30 |
| dansmith | pv_: deploying from git, not from a stable release | 18:31 |
| dansmith | bauzas: okay, we're trying to lift the minimum for that package, but I think we might still be waiting for that to land | 18:31 |
| pv_ | http://pastebin.com/C4syddEN here are the error messages from my cond log | 18:31 |
| *** moshele has joined #openstack-nova | 18:31 | |
| pv_ | yeah im running from master | 18:31 |
| pv_ | should I not be? | 18:31 |
| bauzas | yeah, that seems to be the same problem | 18:31 |
| bauzas | dansmith: sure, I know about rlrossit's change | 18:32 |
| *** sachin has quit IRC | 18:32 | |
| *** sachin has joined #openstack-nova | 18:32 | |
| bauzas | just wondering if pv_'s issue could be related to that - provided he's running from source :) | 18:32 |
| dansmith | bauzas: probably | 18:32 |
| dansmith | although I'm not sure why this would have changed | 18:33 |
| dansmith | pv_: what version of oslo.versionedobjects? | 18:33 |
| pv_ | how do i check that? | 18:34 |
| dansmith | pv_: maybe you could tell me more about your environment.. is this a real deployment, devstack, unit tests... ? | 18:34 |
| pv_ | oh sorry im running devstack | 18:35 |
| pv_ | one a single node on which im testing my FC cinder driver | 18:35 |
| bauzas | so yeah, probably related to that | 18:35 |
| bauzas | cd /opt/stack/nova; sudo pip install -r requirements.txt | 18:35 |
| pv_ | on a * | 18:35 |
| dansmith | okay, I just realized we might've broken something with this: https://review.openstack.org/#/c/217382/1/nova/objects/base.py,cm | 18:36 |
| *** mjturek1 has quit IRC | 18:36 | |
| bauzas | dansmith: if latest o.vo, that's fine | 18:36 |
| dansmith | bauzas: no it's not | 18:36 |
| bauzas | dansmith: I just did a venv upgrade and it went fine | 18:36 |
| bauzas | but I'm not within devstack | 18:37 |
| dansmith | right, but that patch may be breaking wire protocol, which means older nodes will be broken | 18:37 |
| bauzas | oh gotcha ! | 18:37 |
| *** tonytan4ever has quit IRC | 18:38 | |
| *** tonytan4ever has joined #openstack-nova | 18:38 | |
| *** tonytan4ever has quit IRC | 18:38 | |
| *** salv-orlando has joined #openstack-nova | 18:39 | |
| pv_ | should i change it from the master branch to stable-kilo? | 18:40 |
| pv_ | the nova_branch | 18:40 |
| dansmith | oh, yeah, you're running old stuff I think | 18:40 |
| dansmith | so I think this is actually okay because of thangp's patch | 18:40 |
| pv_ | alright ill try that thanks | 18:41 |
| dansmith | pv_: so, just to be sure, please "pip freeze | grep versionedobjects" | 18:41 |
| dansmith | bauzas: so that patch didn't break it as long as you're running a newer o.vo that honors a field | 18:41 |
| pv_ | oslo.versionedobjects==0.7.0 | 18:41 |
| *** takedakn has quit IRC | 18:41 | |
| dansmith | bauzas: which is why it fixes it for you, and why the gate is working I think | 18:41 |
| pv_ | is what im getting from that | 18:42 |
| dansmith | pv_: yeah, so if you could try just upgrading that and re-trying that would be a good test | 18:42 |
| dansmith | pip install -U oslo.versionedobjects | 18:42 |
| dansmith | should get you 0.9.0 | 18:42 |
| pv_ | ok thanks, ill try that | 18:42 |
| bauzas | dansmith: yeah, since the gate is always getting the latest dependencies, we're good | 18:42 |
| bauzas | dansmith: but some people who would just use devstack with an OFFLINE=True could get that ^ | 18:42 |
| *** lpetrut has quit IRC | 18:43 | |
| dansmith | bauzas: by gate, I meant the partial-ncpu job, which should be ensuring we didn't break the wire protocol between K and L | 18:43 |
| *** mjturek1 has joined #openstack-nova | 18:43 | |
| dansmith | bauzas: until we land the patches to bump the minimum, yeah | 18:43 |
| bauzas | dansmith: oh right, that's the server which needs to have the most up-to-date, right? so, since it's the controller, hence why, nope? | 18:44 |
| openstackgerrit | Matt Riedemann proposed openstack/nova: WIP: Allow admin to reboot instances when syncing power states https://review.openstack.org/218975 | 18:44 |
| dansmith | bauzas: this doesn't actually have anything to do with that it's a controller, just that we're running some new code talking to some old code, to make sure they're still speaking the same language | 18:44 |
| dansmith | (so to speak) | 18:44 |
| bauzas | but if I remind correctly, the partial-ncpu is a master AIO + a Kilo n-cpu, right? | 18:45 |
| bauzas | dansmith: ^ | 18:46 |
| dansmith | yeah | 18:46 |
| bauzas | okay, so since the master is always up-to-date, that's working IIUC | 18:46 |
| dansmith | since the master nodes are pulling a new o.vo, yes | 18:46 |
| *** meghal has quit IRC | 18:47 | |
| dansmith | pv_: I do actually want to hear back, FYI :) | 18:47 |
| *** meghal has joined #openstack-nova | 18:47 | |
| pv_ | yeah definitely, im running stack.sh so it'll be a bit but ill lyk! | 18:47 |
| dansmith | pv_: did you get 0.9.0 after the install -U? | 18:47 |
| pv_ | yup | 18:48 |
| dansmith | okay cool | 18:48 |
| pv_ | it went from 0.7.0 to 0.9.0 | 18:48 |
| dansmith | yeah, cool | 18:48 |
| dansmith | the patch to raise the minimum is in the gate now | 18:48 |
| dansmith | that would have fixed this for you | 18:48 |
| *** ig0r__ has quit IRC | 18:48 | |
| *** SourabhP has quit IRC | 18:50 | |
| *** mjturek2 has joined #openstack-nova | 18:52 | |
| *** mjturek1 has quit IRC | 18:53 | |
| *** edmondsw has quit IRC | 18:56 | |
| pv_ | hey thanks so much, it works now the update of oslo.versionedobjects did the trick! | 18:56 |
| *** hemna has joined #openstack-nova | 18:57 | |
| *** edmondsw has joined #openstack-nova | 18:57 | |
| *** ociuhandu has quit IRC | 18:58 | |
| dansmith | pv_: cool | 19:00 |
| *** SourabhP has joined #openstack-nova | 19:01 | |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Don't query database with an empty list of tags for creation https://review.openstack.org/214189 | 19:03 |
| rlrossit | dansmith: the 0.9.0 change just got merged | 19:05 |
| dansmith | rlrossit: nice | 19:06 |
| *** Apoorva has quit IRC | 19:06 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/216968 | 19:07 |
| rlrossit | I want this ^ | 19:10 |
| bauzas | dansmith: since the above is merged, I guess there is nothing to be done ? | 19:11 |
| bauzas | dansmith: or do you plan some defensive approach ? | 19:11 |
| dansmith | bauzas: nope, we're good I think | 19:12 |
| *** jwcroppe has joined #openstack-nova | 19:13 | |
| bauzas | dansmith: that's my understanding | 19:13 |
| *** ZZelle has joined #openstack-nova | 19:13 | |
| *** mjturek1 has joined #openstack-nova | 19:14 | |
| *** jwcroppe_ has quit IRC | 19:15 | |
| *** mjturek2 has quit IRC | 19:15 | |
| *** tojuvone_ has quit IRC | 19:18 | |
| *** Apoorva has joined #openstack-nova | 19:21 | |
| *** pixelb has quit IRC | 19:23 | |
| *** baoli has quit IRC | 19:24 | |
| *** baoli has joined #openstack-nova | 19:24 | |
| *** lpetrut has joined #openstack-nova | 19:29 | |
| *** promulo__ has quit IRC | 19:32 | |
| *** exploreshaifali has joined #openstack-nova | 19:33 | |
| *** davideag_ has joined #openstack-nova | 19:35 | |
| *** davideagnello has quit IRC | 19:37 | |
| *** salv-orlando has quit IRC | 19:37 | |
| *** jichen has joined #openstack-nova | 19:42 | |
| *** thorst1 has quit IRC | 19:47 | |
| bauzas | dansmith: got a few mins for just answering a question ? | 19:49 |
| bauzas | dansmith: why is there an expected_attrs param for _from_db_object() if we provide a obj_lazy_load() method for those fields ? | 19:49 |
| dansmith | bauzas: in the middle of something, maybe in a bit? | 19:49 |
| *** Sukhdev has joined #openstack-nova | 19:49 | |
| bauzas | dansmith: ack, no worries | 19:49 |
| dansmith | bauzas: because we always want to load them if they're there | 19:50 |
| bauzas | dansmith: I'm just trying to understand some logic, not rush | 19:50 |
| *** salv-orlando has joined #openstack-nova | 19:50 | |
| dansmith | bauzas: if we asked for them in the db query, we need to tell from_db_obj to load them | 19:50 |
| dansmith | however, if we didn't we don't want it to try to load them, because that would lazy-load from the database, creating more hits, | 19:50 |
| dansmith | and also always return everything from the rpc calls, which we don't always want | 19:50 |
| bauzas | dansmith: I see, it provides a way to either get them or not | 19:51 |
| bauzas | and if the latter, then lazy-load it | 19:51 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: RT: Migration resource tracking uses migration context https://review.openstack.org/218938 | 19:51 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: compute: migrate/resize paths properly handle stashed numa_topology https://review.openstack.org/218500 | 19:51 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: Claims: Make sure move claims crate a migration context records https://review.openstack.org/218385 | 19:51 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: objects: Hook migration object into Instance https://review.openstack.org/216733 | 19:51 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: objects: Add MigrationContext object https://review.openstack.org/216732 | 19:51 |
| ndipanov | dansmith, pls^ | 19:51 |
| bauzas | dansmith: gotcha, thanks | 19:51 |
| ndipanov | will heck back in a bit if there is some nits | 19:51 |
| dansmith | ndipanov: so you still have the get-by-uuid | 19:52 |
| dansmith | you need that? | 19:52 |
| *** davideagnello has joined #openstack-nova | 19:52 | |
| *** VW__ has joined #openstack-nova | 19:52 | |
| ndipanov | yeah for the lazy load no | 19:52 |
| dansmith | the _save shouldn't be necessary, as you can just set that in the instance save itself and do it in one shot | 19:52 |
| dansmith | ndipanov: ah, okay | 19:53 |
| ndipanov | it's not a remotable so we can move it if we don't like it (_save) | 19:53 |
| dansmith | yep | 19:53 |
| *** davideag_ has quit IRC | 19:53 | |
| dansmith | I think we should, but since it's changeable later, that's fine for now | 19:53 |
| *** VW__ has quit IRC | 19:55 | |
| dansmith | ndipanov: why not make apply and revert remotable and atomic in terms of unsetting the context? | 19:55 |
| *** VW__ has joined #openstack-nova | 19:55 | |
| *** VW_ has quit IRC | 19:56 | |
| dansmith | okay, I see | 19:57 |
| *** rushil has joined #openstack-nova | 19:58 | |
| *** mspreitz has joined #openstack-nova | 19:58 | |
| dansmith | jaypipes: I made ndipanov change a few things around the object methods, so if you could re-review and +W please: https://review.openstack.org/#/c/216732/5 | 19:59 |
| dansmith | jaypipes: very little change, just a little more efficiency in how we embed the objects | 19:59 |
| *** VW__ has quit IRC | 19:59 | |
| *** VW_ has joined #openstack-nova | 20:00 | |
| *** david-lyle has quit IRC | 20:01 | |
| jaypipes | dansmith: done | 20:02 |
| dansmith | thanks | 20:02 |
| jaypipes | np | 20:02 |
| *** electrocucaracha has quit IRC | 20:02 | |
| *** lbeliveau has joined #openstack-nova | 20:03 | |
| dansmith | jaypipes: there are also patches above that that you previously +Wed ... :) | 20:05 |
| *** markvoelker has joined #openstack-nova | 20:06 | |
| *** david-lyle has joined #openstack-nova | 20:06 | |
| *** electrocucaracha has joined #openstack-nova | 20:06 | |
| openstackgerrit | jichenjc proposed openstack/nova: API: remove unused parameter https://review.openstack.org/218932 | 20:06 |
| mriedem | garyk: could use some vmware reviews on this https://review.openstack.org/#/c/183985/ | 20:11 |
| mriedem | also looked for browne and tjones but they aren't around | 20:11 |
| *** thedodd has quit IRC | 20:11 | |
| *** dims has joined #openstack-nova | 20:12 | |
| *** dims has quit IRC | 20:12 | |
| *** VW_ has quit IRC | 20:12 | |
| *** VW_ has joined #openstack-nova | 20:12 | |
| *** ivasev has quit IRC | 20:13 | |
| *** thorst1 has joined #openstack-nova | 20:14 | |
| *** dims_ has quit IRC | 20:14 | |
| pv_ | had another question, im currently debugging my volume-attach functionality in my cinder driver | 20:15 |
| pv_ | its an FC driver and the machine im running devstack on is definitely HBA enabled | 20:15 |
| pv_ | but im getting this in the n-cpu logs | 20:15 |
| pv_ | http://pastebin.com/iNs2bT07 | 20:15 |
| *** dprince has quit IRC | 20:16 | |
| pv_ | after my c-vol log says "initialize connection completed succecssfully" | 20:16 |
| pv_ | when i run systool -c fc_host -v im getting both the WWNs that are on the machine but nova isn't getting them according to the n-cpu log | 20:16 |
| pv_ | any help? | 20:16 |
| *** meghal has quit IRC | 20:20 | |
| *** rushil has quit IRC | 20:20 | |
| *** Apoorva has quit IRC | 20:22 | |
| mriedem | pv_: do you have the latest os-brick release? | 20:24 |
| mriedem | 0.3.2 | 20:24 |
| mriedem | pv_: anyway, in liberty, nova proxies to os-brick for FC stuff | 20:24 |
| pv_ | it seems so | 20:24 |
| pv_ | pip freeze | grep os-brick gives me 0.3.2 | 20:24 |
| mriedem | so it's a better question for the openstack-cinder channel | 20:24 |
| pv_ | yeah for sure, i asked there to | 20:24 |
| pv_ | thansk | 20:24 |
| pv_ | too* | 20:24 |
| mriedem | nova master plan is to proxy all things to other services so we can also proxy the support questions | 20:25 |
| *** Apoorva has joined #openstack-nova | 20:25 | |
| *** thangp has quit IRC | 20:32 | |
| *** matrohon has joined #openstack-nova | 20:33 | |
| *** diogogmt has joined #openstack-nova | 20:36 | |
| *** lpetrut has quit IRC | 20:36 | |
| *** rlrossit1 has joined #openstack-nova | 20:37 | |
| *** achanda has quit IRC | 20:38 | |
| *** rlrossit has quit IRC | 20:39 | |
| *** sachin has quit IRC | 20:40 | |
| *** VW_ has quit IRC | 20:41 | |
| *** sachin has joined #openstack-nova | 20:41 | |
| *** achanda has joined #openstack-nova | 20:41 | |
| *** achanda has quit IRC | 20:41 | |
| *** VW_ has joined #openstack-nova | 20:41 | |
| *** achanda has joined #openstack-nova | 20:41 | |
| *** dprince has joined #openstack-nova | 20:42 | |
| *** electrocucaracha has quit IRC | 20:43 | |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: RT: Migration resource tracking uses migration context https://review.openstack.org/218938 | 20:44 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: compute: migrate/resize paths properly handle stashed numa_topology https://review.openstack.org/218500 | 20:44 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: Claims: Make sure move claims create a migration context records https://review.openstack.org/218385 | 20:44 |
| *** moshele has quit IRC | 20:45 | |
| openstackgerrit | Andrew Laski proposed openstack/nova: WIP Cells: update block device mappings synchronously https://review.openstack.org/219022 | 20:45 |
| *** VW_ has quit IRC | 20:46 | |
| *** rushil has joined #openstack-nova | 20:46 | |
| *** electrocucaracha has joined #openstack-nova | 20:46 | |
| *** bkopilov has quit IRC | 20:48 | |
| *** VW_ has joined #openstack-nova | 20:50 | |
| *** baoli has quit IRC | 20:51 | |
| *** bkopilov has joined #openstack-nova | 20:53 | |
| *** ildikov has quit IRC | 20:54 | |
| *** xyang1 has quit IRC | 20:55 | |
| *** yamahata has quit IRC | 20:55 | |
| *** jwcroppe_ has joined #openstack-nova | 20:55 | |
| *** mwagner_lap has quit IRC | 20:56 | |
| *** markvoelker has quit IRC | 20:56 | |
| *** jwcroppe has quit IRC | 20:57 | |
| *** rlrossit1 has left #openstack-nova | 20:57 | |
| *** rfolco has quit IRC | 20:57 | |
| *** yamahata has joined #openstack-nova | 20:57 | |
| *** dims has joined #openstack-nova | 20:57 | |
| *** Apoorva_ has joined #openstack-nova | 21:00 | |
| *** ccarmack has left #openstack-nova | 21:00 | |
| *** bkopilov has quit IRC | 21:00 | |
| *** xyang1 has joined #openstack-nova | 21:01 | |
| *** jwcroppe has joined #openstack-nova | 21:02 | |
| *** sachin has quit IRC | 21:02 | |
| *** jwcroppe_ has quit IRC | 21:04 | |
| *** Apoorva has quit IRC | 21:04 | |
| *** annegentle has quit IRC | 21:04 | |
| *** jichen has quit IRC | 21:05 | |
| *** jwcroppe_ has joined #openstack-nova | 21:07 | |
| *** jwcroppe has quit IRC | 21:10 | |
| *** hemna has quit IRC | 21:10 | |
| bauzas | jaypipes: so, since https://review.openstack.org/#/c/191250/27 is still in the gate for the next hour, I have my local branch which fixes the merge conflict and provides the new InstanceGroup object waiting to be sent | 21:10 |
| bauzas | jaypipes: I will probably send the whole series by tomorrow morning | 21:11 |
| bauzas | that's basically a FYI | 21:11 |
| bauzas | ::) | 21:11 |
| openstackgerrit | Cedric Brandily proposed openstack/python-novaclient: Allow to reboot multiple servers https://review.openstack.org/218231 | 21:11 |
| *** meghal has joined #openstack-nova | 21:12 | |
| *** dprince has quit IRC | 21:18 | |
| *** mspreitz has quit IRC | 21:19 | |
| *** bkopilov has joined #openstack-nova | 21:20 | |
| *** dustins has quit IRC | 21:26 | |
| *** burgerk has quit IRC | 21:26 | |
| *** jwcroppe has joined #openstack-nova | 21:28 | |
| *** zzzeek has quit IRC | 21:29 | |
| *** jwcroppe_ has quit IRC | 21:31 | |
| *** sachin has joined #openstack-nova | 21:32 | |
| *** annegentle has joined #openstack-nova | 21:35 | |
| *** hemna has joined #openstack-nova | 21:36 | |
| *** annegent_ has joined #openstack-nova | 21:38 | |
| *** matrohon has quit IRC | 21:39 | |
| *** annegentle has quit IRC | 21:40 | |
| *** claudiub has quit IRC | 21:40 | |
| *** Guest15619 is now known as mcarden | 21:45 | |
| *** jcoufal has quit IRC | 21:45 | |
| *** dave-mccowan has quit IRC | 21:46 | |
| *** Sukhdev has quit IRC | 21:49 | |
| *** Sukhdev has joined #openstack-nova | 21:51 | |
| *** IanGovett has quit IRC | 21:57 | |
| *** VW__ has joined #openstack-nova | 21:58 | |
| *** VW_ has quit IRC | 21:58 | |
| *** VW__ has quit IRC | 22:02 | |
| *** meghal has quit IRC | 22:02 | |
| *** meghal has joined #openstack-nova | 22:02 | |
| *** VW_ has joined #openstack-nova | 22:03 | |
| *** meghal has quit IRC | 22:03 | |
| *** meghal has joined #openstack-nova | 22:03 | |
| *** vladikr has quit IRC | 22:04 | |
| *** sachin has quit IRC | 22:04 | |
| *** thedodd has joined #openstack-nova | 22:05 | |
| *** sachin has joined #openstack-nova | 22:06 | |
| *** meghal has quit IRC | 22:06 | |
| *** meghal has joined #openstack-nova | 22:06 | |
| *** mdrabe has quit IRC | 22:06 | |
| *** zzzeek has joined #openstack-nova | 22:06 | |
| *** eharney has quit IRC | 22:07 | |
| *** sachin has quit IRC | 22:07 | |
| *** sachin has joined #openstack-nova | 22:08 | |
| *** sachin has quit IRC | 22:08 | |
| *** sachin_ has joined #openstack-nova | 22:09 | |
| *** sachin_ has quit IRC | 22:10 | |
| *** sachin has joined #openstack-nova | 22:10 | |
| *** sachin has quit IRC | 22:11 | |
| *** markvoelker has joined #openstack-nova | 22:11 | |
| *** sachin has joined #openstack-nova | 22:11 | |
| mriedem | dansmith: when you get a chance, could use your thoughts on only waiting for vif unplug events during rebuild here https://review.openstack.org/#/c/210126/8/nova/virt/libvirt/driver.py | 22:15 |
| mriedem | i can't think of a reason why we'd care about waiting for the ports to be unplugged if we're just deleting the instance (forever) | 22:16 |
| *** markvoelker has quit IRC | 22:16 | |
| dansmith | mriedem: well, I would think we'd want some input from other driver writers | 22:16 |
| dansmith | mriedem: there may be some situations where a hypervisor can't unplug if it's still in use, right? | 22:16 |
| mriedem | this change is only in the libvirt driver | 22:17 |
| mriedem | and i think the libvirt driver is the only one that implements the vif-plugged event wait callback stuf | 22:17 |
| mriedem | *stuff | 22:17 |
| dansmith | well, I meant hypervisor and vif drivers, fwiw | 22:17 |
| mriedem | hrm | 22:18 |
| *** asettle has joined #openstack-nova | 22:18 | |
| mriedem | mellanox giving me vif unplug problems | 22:18 |
| mriedem | still, in this change it's configurable, just like the vif plugging stuff is configurable, | 22:18 |
| mriedem | but i asked him to scope it only to the rebuild task | 22:18 |
| dansmith | so, we don't wait for these now, | 22:20 |
| dansmith | you're saying you think we *should* be waiting for them in the rebuild case, but only that case? | 22:21 |
| mriedem | i'm saying if we should, i tihnk we should restrict it to rebuild | 22:22 |
| *** meghal has quit IRC | 22:22 | |
| mriedem | i don't actually have a recreate of the bug | 22:22 |
| *** Piet has quit IRC | 22:22 | |
| mikal | mriedem: you got a sec to chat about a review you did? | 22:22 |
| mriedem | i asked them why the rebuild test wasn't failing with tempest in the gate and they said it's a race when you have lots of ports | 22:22 |
| mriedem | mikal: am i in trouble? | 22:22 |
| *** meghal has joined #openstack-nova | 22:22 | |
| mriedem | beyond f'ing the gate this afternoon? | 22:22 |
| mikal | mriedem: no, I just want to work out if I agree with you or not | 22:22 |
| mikal | mriedem: https://review.openstack.org/#/c/215414 | 22:22 |
| mriedem | then ok | 22:22 |
| mikal | mriedem: I'm not sure that is a feature -- or if it is its super trivial | 22:23 |
| mikal | mriedem: it feels to me more like an error of omission than a feature being added | 22:23 |
| mriedem | mikal: i saw config option to enable support for a thing | 22:23 |
| mriedem | that smells like a feature ot me | 22:23 |
| mriedem | *to | 22:23 |
| *** sachin has quit IRC | 22:23 | |
| mriedem | i told them specless blueprint maybe | 22:24 |
| mriedem | but still seemed like a blueprint was required | 22:24 |
| mikal | mriedem: but there's no design right? Its just a flag. | 22:24 |
| mikal | mriedem: we have bug fixes add flags all the time. | 22:24 |
| mriedem | mikal: this was the other https://review.openstack.org/#/c/185129/ | 22:24 |
| * mikal looks | 22:24 | |
| mriedem | mikal: i'm all for considering it 'stuck' and discussing it in the nova meeting | 22:25 |
| dansmith | mriedem: commented on how that patch is fundamentally missing the point | 22:25 |
| mriedem | dansmith: aight | 22:25 |
| mikal | Oh, so they tried exactly the same change before, it got blocked asking for a bp review, so they just made it a bug? | 22:25 |
| mriedem | mikal: no, coincidental | 22:25 |
| mikal | Is the image metadata entry new too? | 22:25 |
| mriedem | different teams i think | 22:25 |
| mikal | Ahhh, ok | 22:25 |
| mriedem | mikal: i guess in the first patch they add the metadata to the image to enable the thing in nova, | 22:26 |
| dansmith | mikal: yeah, that aarch thing needs design | 22:26 |
| mriedem | in the newer patch they add a config option for that | 22:26 |
| *** meghal has quit IRC | 22:27 | |
| *** meghal1 has joined #openstack-nova | 22:27 | |
| dansmith | this is something we need to get right, and we've had some arm folks thinking about how to properly represent it in glance | 22:27 |
| dansmith | and it's non-trivial from what I understand | 22:27 |
| mikal | dansmith: why is it hard? Is the 10 line patch super naive? | 22:27 |
| *** sachin has joined #openstack-nova | 22:27 | |
| * mikal knows bugger all about arm | 22:27 | |
| dansmith | mikal: yeah, I think it is.. a given image needs to have an nvram and pram (or something) component with it, kinda like uec images are three pieces | 22:27 |
| dansmith | and in order for us do it right, nova needs to clone the nvram and customize it for the individual instance or some such | 22:28 |
| mikal | dansmith: huh, fair enough | 22:28 |
| *** edmondsw has quit IRC | 22:28 | |
| mikal | Ok, fair enough. I figured I should just ask for an explaination given I wandered across that review and was confused. Thanks. | 22:29 |
| mikal | Sounds like something we should consider for the summit though. | 22:29 |
| * mriedem feels justified in the -2 | 22:29 | |
| dansmith | rich jones has done a bunch of arm stuff and has been doing some thinking on the glance thing | 22:29 |
| dansmith | not sure if he'll be at summit, but I hope so | 22:29 |
| dansmith | I was talking with him about it at kvm forum a couple weeks ago | 22:30 |
| mikal | I also don't know if this is worth an entire summit session -- I don't grok the scope of the problem enough to know. | 22:30 |
| mriedem | certainly not a summit session imo | 22:31 |
| mriedem | we have bigger fish to fry right? | 22:32 |
| *** jaypipes has quit IRC | 22:32 | |
| *** annegent_ has quit IRC | 22:32 | |
| mriedem | i mean, you all have bigger fish to fry | 22:32 |
| *** annegentle has joined #openstack-nova | 22:32 | |
| mikal | mriedem: depends on how complicated it is I guess | 22:33 |
| mikal | mriedem: if it requires major replumbing... | 22:33 |
| *** ociuhandu has joined #openstack-nova | 22:33 | |
| *** thorst1 has quit IRC | 22:35 | |
| *** mjturek1 has quit IRC | 22:35 | |
| *** meghal1 has quit IRC | 22:36 | |
| *** meghal has joined #openstack-nova | 22:36 | |
| *** hemna has quit IRC | 22:38 | |
| *** ijuwang has quit IRC | 22:40 | |
| *** ctrath has quit IRC | 22:42 | |
| mikal | mriedem: I kind of want you to say something on https://review.openstack.org/#/c/214773/ when you have a moment | 22:47 |
| mikal | mriedem: am I wrong to think that's a back port of a feature? | 22:48 |
| mriedem | mikal: of the 3 bugs listed, https://bugs.launchpad.net/nova/+bug/1246201 is the only one that says anything about live migration | 22:49 |
| openstack | Launchpad bug 1246201 in OpenStack Compute (nova) "Live migration fails when the instance has a config_drive_format=iso9660" [High,In progress] - Assigned to Pavel Boldin (pboldin) | 22:49 |
| mikal | mriedem: I think what dreamhost is saying is that the only form of live migration they want to support is via ceph | 22:50 |
| mikal | mriedem: I suggested they could block migrate the config drive, and they were ... unenthused | 22:50 |
| *** VW__ has joined #openstack-nova | 22:50 | |
| *** VW_ has quit IRC | 22:51 | |
| *** VW__ has quit IRC | 22:51 | |
| *** VW_ has joined #openstack-nova | 22:51 | |
| mriedem | idk, i'd have to dig into the change history on the bugs and figure out wtf is going on, it looks like the be all end all fix for people using ceph | 22:51 |
| clarkb | isnt it small enough to write and attach a new one? | 22:51 |
| *** meghal has quit IRC | 22:51 | |
| mikal | clarkb: we can't re-create the config drive post boot | 22:52 |
| *** Piet has joined #openstack-nova | 22:52 | |
| mikal | clarkb: some of the data is gone at that point | 22:52 |
| clarkb | ah | 22:52 |
| mikal | clarkb: we _have_ to copy it somehow | 22:52 |
| mikal | mriedem: so, I guess my argument is this -- there was no attempt to store config drives on ceph in kilo. This adds that. Adding that is a feature not a bug. | 22:52 |
| mikal | mriedem: and sure, having it not in kilo might hurt for some people, but that pain isn't new or unexpected. | 22:53 |
| mriedem | mikal: heh, i just realized they were cherry picking your fix | 22:55 |
| mikal | mriedem: its true. I am arguing that we shouldn't merge code from me. | 22:55 |
| mikal | mriedem: I am ... conflicted on that point :P | 22:55 |
| mriedem | mikal: https://review.openstack.org/#/c/116847/ and https://review.openstack.org/#/c/112014/ were in stable/kilo | 22:56 |
| *** VW_ has quit IRC | 22:56 | |
| mriedem | then reverted at some point ? | 22:56 |
| *** VW_ has joined #openstack-nova | 22:56 | |
| mikal | mriedem: I'd had to dig | 22:56 |
| *** VW_ has quit IRC | 22:57 | |
| mikal | mriedem: we never shipped ceph support for config drive because it was broken and fast reverted | 22:57 |
| *** VW_ has joined #openstack-nova | 22:57 | |
| mikal | mriedem: this is the patch I wrote as an apology for reverting the broken thing | 22:57 |
| mriedem | ah | 22:57 |
| *** VW_ has quit IRC | 22:58 | |
| mriedem | so if this adds config drive support for rbd to kilo, then i agree it's probably not appropriate. i feel like it's trying to come in under the guise of being for live migration to work | 22:58 |
| mriedem | i guess it's, well live migration would work with ceph + config drive if yo uland this | 22:58 |
| mikal | Yeah, the details of the reverts are in the commit message for the review that started this conversation | 22:58 |
| *** VW_ has joined #openstack-nova | 22:58 | |
| mikal | mriedem: correct, I believe live migration works for these users if they use block migration | 22:58 |
| mikal | mriedem: if they insist on only using ceph shared storage migration, they have snookered themselves | 22:59 |
| mikal | mriedem: yeah, those two reviews you found are reverted | 22:59 |
| mikal | mriedem: and were never in an official release (IIRC) | 23:00 |
| *** mwagner_lap has joined #openstack-nova | 23:00 | |
| mriedem | if this were juno i'd definitely be -1 on a thing like this just given the size of the change | 23:01 |
| mikal | mriedem: https://review.openstack.org/#/c/123070/ is the revert | 23:01 |
| mriedem | ok | 23:01 |
| mriedem | well, | 23:02 |
| mikal | mriedem: and https://review.openstack.org/#/c/123071/ | 23:02 |
| mriedem | i'm at least on the -1 side just due to complexity | 23:02 |
| mriedem | and lack of ceph testing in kilo | 23:02 |
| *** VW_ has quit IRC | 23:02 | |
| mikal | mriedem: and https://review.openstack.org/#/c/123072/ | 23:02 |
| mriedem | oh i guess there is the ceph job on stable/kilo huh | 23:02 |
| mriedem | we don't test live migration + ceph but ok | 23:03 |
| * mikal wishes gerrit would make it more obvious that a review was later reverted | 23:03 | |
| mriedem | mikal: it does if you revert via the button in gerrit | 23:03 |
| mikal | mriedem: oh, that's news to me | 23:03 |
| mriedem | yeah the ceph job passed on that change http://logs.openstack.org/73/214773/1/check/gate-tempest-dsvm-full-ceph/43f6f62/ | 23:04 |
| *** annegentle has quit IRC | 23:04 | |
| *** exploreshaifali has quit IRC | 23:04 | |
| *** lucas-dinner has quit IRC | 23:05 | |
| mriedem | mikal: there, -1 with my most eloquent apology possible right now | 23:07 |
| *** apuimedo|away has quit IRC | 23:07 | |
| mikal | mriedem: heh, thanks. I think. | 23:07 |
| *** thedodd has quit IRC | 23:12 | |
| *** irenab_ has joined #openstack-nova | 23:14 | |
| *** irenab has quit IRC | 23:16 | |
| *** tanee has quit IRC | 23:16 | |
| *** irenab_ is now known as irenab | 23:16 | |
| *** tanee has joined #openstack-nova | 23:16 | |
| *** Apoorva has joined #openstack-nova | 23:17 | |
| *** Apoorva has quit IRC | 23:17 | |
| *** Apoorva has joined #openstack-nova | 23:18 | |
| *** Apoorva_ has quit IRC | 23:20 | |
| *** sachin has quit IRC | 23:22 | |
| *** meghal has joined #openstack-nova | 23:23 | |
| *** sachin has joined #openstack-nova | 23:23 | |
| *** rushil has quit IRC | 23:24 | |
| *** shyu has joined #openstack-nova | 23:24 | |
| *** rushil has joined #openstack-nova | 23:32 | |
| *** ZZelle has quit IRC | 23:35 | |
| *** rushil has quit IRC | 23:36 | |
| *** rushil has joined #openstack-nova | 23:36 | |
| *** dave-mccowan has joined #openstack-nova | 23:41 | |
| cfriesen | is there a known issue with gate-grenade-dsvm ? http://logs.openstack.org/89/210689/4/check/gate-grenade-dsvm/e90059f/logs/grenade.sh.txt.gz seems to show it complaining about trying to use pbr 0.11.0 which doesn't meet requirements. | 23:43 |
| *** emagana has quit IRC | 23:44 | |
| *** shoutm has joined #openstack-nova | 23:46 | |
| *** electrocucaracha has quit IRC | 23:50 | |
| *** Sukhdev has quit IRC | 23:51 | |
| *** diegows has quit IRC | 23:53 | |
| *** andreykurilin__ has quit IRC | 23:55 | |
| *** achanda_ has joined #openstack-nova | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!