| *** Raildo_ has joined #openstack-nova | 00:05 | |
| *** markvoelker has joined #openstack-nova | 00:06 | |
| *** achanda_ has joined #openstack-nova | 00:06 | |
| *** 92AAA7LGD has quit IRC | 00:09 | |
| *** 18WAABPY2 has quit IRC | 00:09 | |
| *** Marga_ has quit IRC | 00:10 | |
| *** achanda has quit IRC | 00:10 | |
| *** Marga_ has joined #openstack-nova | 00:10 | |
| *** achanda_ has quit IRC | 00:11 | |
| *** derekh has quit IRC | 00:12 | |
| *** salv-orlando has joined #openstack-nova | 00:15 | |
| *** annegentle has joined #openstack-nova | 00:17 | |
| *** otter768 has joined #openstack-nova | 00:21 | |
| *** garyk has joined #openstack-nova | 00:21 | |
| openstackgerrit | sajeesh cs proposed openstack/nova: Create column allocated in Quota table https://review.openstack.org/151327 | 00:22 |
|---|---|---|
| *** annegentle has quit IRC | 00:23 | |
| *** yingjun has joined #openstack-nova | 00:23 | |
| *** tsekiyam_ has joined #openstack-nova | 00:25 | |
| *** otter768 has quit IRC | 00:25 | |
| *** zsais has quit IRC | 00:25 | |
| *** salv-orlando has quit IRC | 00:26 | |
| *** promulo_ has joined #openstack-nova | 00:26 | |
| *** mtanino has quit IRC | 00:27 | |
| vilobhmm1 | johnthetubaguy : https://review.openstack.org/#/c/138607/ whenever you are free please let me know if looks ok to you… | 00:28 |
| *** david-lyle has quit IRC | 00:28 | |
| vilobhmm1 | and good to remove the X | 00:28 |
| *** tsekiyama has quit IRC | 00:28 | |
| *** tsekiyam_ has quit IRC | 00:29 | |
| *** promulo__ has quit IRC | 00:30 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Ignore me. trying to log where urllib3 connectionpool is created https://review.openstack.org/182495 | 00:33 |
| *** annegentle has joined #openstack-nova | 00:33 | |
| *** IanGovett has quit IRC | 00:40 | |
| jogo | wee gate is blocked | 00:41 |
| *** angdraug has quit IRC | 00:41 | |
| *** marun has quit IRC | 00:48 | |
| *** jdurgin has quit IRC | 00:50 | |
| *** browne has quit IRC | 00:53 | |
| openstackgerrit | sajeesh cs proposed openstack/nova: Create column allocated in Quota table https://review.openstack.org/151327 | 00:53 |
| *** rwsu has quit IRC | 00:54 | |
| *** Raildo__ has joined #openstack-nova | 00:54 | |
| *** ferest has joined #openstack-nova | 00:57 | |
| *** Raildo_ has quit IRC | 00:58 | |
| *** erickson has joined #openstack-nova | 00:59 | |
| *** ericksonfgds has quit IRC | 01:00 | |
| *** Nic has quit IRC | 01:01 | |
| *** oro has joined #openstack-nova | 01:01 | |
| openstackgerrit | sajeesh cs proposed openstack/nova: Setting the default values of resources to zero. https://review.openstack.org/151677 | 01:03 |
| *** baoli has quit IRC | 01:10 | |
| *** oro has quit IRC | 01:10 | |
| *** signed8bit_ZZZzz is now known as signed8b_ | 01:12 | |
| *** claudiub|2 has quit IRC | 01:12 | |
| openstackgerrit | sajeesh cs proposed openstack/nova: Removal of redundant context checking from wsgi.py ,which is common to all api calls https://review.openstack.org/182140 | 01:13 |
| *** dave-mccowan has quit IRC | 01:20 | |
| *** alanf-mc has quit IRC | 01:20 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Ignore me. trying to log where urllib3 connectionpool is created https://review.openstack.org/182495 | 01:21 |
| *** marun has joined #openstack-nova | 01:24 | |
| *** popw has joined #openstack-nova | 01:26 | |
| *** claudiub has joined #openstack-nova | 01:28 | |
| *** VW_ has joined #openstack-nova | 01:30 | |
| *** VW_ has quit IRC | 01:30 | |
| *** VW_ has joined #openstack-nova | 01:31 | |
| *** Sukhdev has joined #openstack-nova | 01:33 | |
| *** Raildo_ has joined #openstack-nova | 01:35 | |
| *** oro has joined #openstack-nova | 01:35 | |
| *** oro_ has joined #openstack-nova | 01:35 | |
| *** tobe43 has joined #openstack-nova | 01:38 | |
| *** Raildo__ has quit IRC | 01:39 | |
| *** Raildo_ has quit IRC | 01:39 | |
| *** mestery has quit IRC | 01:45 | |
| *** zzzeek has quit IRC | 01:47 | |
| *** erickson has quit IRC | 01:47 | |
| *** vilobhmm1 has quit IRC | 01:49 | |
| *** vilobhmm1 has joined #openstack-nova | 01:49 | |
| *** Sukhdev_ has joined #openstack-nova | 01:53 | |
| *** ferest has quit IRC | 01:53 | |
| *** vilobhmm1 has quit IRC | 01:54 | |
| *** Sukhdev has quit IRC | 01:54 | |
| *** slong has joined #openstack-nova | 01:56 | |
| *** claudiub has quit IRC | 02:02 | |
| *** patrickeast has quit IRC | 02:03 | |
| *** annegentle has quit IRC | 02:07 | |
| openstackgerrit | garyk proposed openstack/nova: libvirt: improve logging https://review.openstack.org/182452 | 02:10 |
| openstackgerrit | garyk proposed openstack/nova: libvirt: use instance UUID with exception InstanceNotFound https://review.openstack.org/182442 | 02:10 |
| *** erickson has joined #openstack-nova | 02:11 | |
| *** browne has joined #openstack-nova | 02:17 | |
| *** otter768 has joined #openstack-nova | 02:22 | |
| openstackgerrit | sajeesh cs proposed openstack/nova: Two roles ,namely project_admin and cloud_admin are created. For updating or deleting the quota, the user should be either project_admin or cloud_admin or admin But only admin or cloud_admin can update or delete the quota of the root project. https://review.openstack.org/182522 | 02:24 |
| *** zzzeek has joined #openstack-nova | 02:26 | |
| *** zzzeek has quit IRC | 02:26 | |
| *** otter768 has quit IRC | 02:26 | |
| *** pixelbeat_ has quit IRC | 02:27 | |
| *** yingjun has quit IRC | 02:29 | |
| *** dave-mccowan has joined #openstack-nova | 02:30 | |
| *** baoli has joined #openstack-nova | 02:30 | |
| *** yingjun has joined #openstack-nova | 02:32 | |
| *** dave-mccowan has quit IRC | 02:38 | |
| *** yamahata has quit IRC | 02:39 | |
| *** annegentle has joined #openstack-nova | 02:40 | |
| *** annegentle has quit IRC | 02:41 | |
| *** baoli has quit IRC | 02:44 | |
| *** achanda has joined #openstack-nova | 02:47 | |
| *** gyee has quit IRC | 02:47 | |
| *** sdake has quit IRC | 02:51 | |
| *** dsanders has quit IRC | 02:54 | |
| *** oro_ has quit IRC | 02:54 | |
| *** oro has quit IRC | 02:55 | |
| *** unicell has quit IRC | 02:58 | |
| *** tobe43 has quit IRC | 02:58 | |
| *** alexpilotti has quit IRC | 03:00 | |
| *** signed8b_ is now known as signed8bit_ZZZzz | 03:01 | |
| *** baoli has joined #openstack-nova | 03:02 | |
| *** yamahata has joined #openstack-nova | 03:03 | |
| *** erickson has quit IRC | 03:10 | |
| *** annegentle has joined #openstack-nova | 03:10 | |
| *** slong has quit IRC | 03:12 | |
| *** dave-mccowan has joined #openstack-nova | 03:14 | |
| *** dave-mcc_ has joined #openstack-nova | 03:14 | |
| *** signed8bit_ZZZzz is now known as signed8b_ | 03:16 | |
| *** salv-orlando has joined #openstack-nova | 03:17 | |
| *** signed8b_ is now known as signed8bit_ZZZzz | 03:17 | |
| *** dims_ has quit IRC | 03:18 | |
| *** annegentle has quit IRC | 03:20 | |
| *** links has joined #openstack-nova | 03:30 | |
| *** nickchase has quit IRC | 03:32 | |
| *** vilobhmm1 has joined #openstack-nova | 03:33 | |
| *** zhangtralon has joined #openstack-nova | 03:38 | |
| *** zhangtralon has quit IRC | 03:39 | |
| *** yingjun has quit IRC | 03:40 | |
| *** marun has quit IRC | 03:41 | |
| *** patrickeast has joined #openstack-nova | 03:42 | |
| *** salv-orlando has quit IRC | 03:42 | |
| *** Sree_ has joined #openstack-nova | 03:43 | |
| *** Jeffrey4l has quit IRC | 03:43 | |
| *** Sree_ is now known as Guest94770 | 03:43 | |
| *** david-lyle has joined #openstack-nova | 03:48 | |
| *** asettle has joined #openstack-nova | 03:49 | |
| *** baoli has quit IRC | 03:49 | |
| *** zhangtralon has joined #openstack-nova | 03:49 | |
| *** oro has joined #openstack-nova | 03:52 | |
| *** oro_ has joined #openstack-nova | 03:52 | |
| *** Sukhdev_ has quit IRC | 03:53 | |
| *** dave-mccowan has quit IRC | 03:53 | |
| *** dave-mcc_ has quit IRC | 03:53 | |
| *** Jeffrey4l has joined #openstack-nova | 03:56 | |
| *** coolsvap|afk is now known as coolsvap | 03:56 | |
| *** deepthi has joined #openstack-nova | 04:04 | |
| *** salv-orlando has joined #openstack-nova | 04:10 | |
| *** dims has joined #openstack-nova | 04:18 | |
| *** otter768 has joined #openstack-nova | 04:22 | |
| *** ildikov has quit IRC | 04:23 | |
| *** salv-orlando has quit IRC | 04:25 | |
| *** dims has quit IRC | 04:25 | |
| *** zhangtralon has quit IRC | 04:25 | |
| *** vilobhmm1 has quit IRC | 04:25 | |
| *** vilobhmm1 has joined #openstack-nova | 04:26 | |
| *** otter768 has quit IRC | 04:27 | |
| *** tojuvone has joined #openstack-nova | 04:29 | |
| *** salv-orlando has joined #openstack-nova | 04:32 | |
| *** salv-orlando has quit IRC | 04:33 | |
| *** Guest94770 has quit IRC | 04:34 | |
| *** ildikov has joined #openstack-nova | 04:36 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Use host AZ instead of instance AZ in scheduler request_spec https://review.openstack.org/174269 | 04:37 |
| *** yingjun has joined #openstack-nova | 04:41 | |
| *** deepakcs has joined #openstack-nova | 04:41 | |
| deepakcs | johnthetubaguy: ping | 04:41 |
| *** ildikov has quit IRC | 04:43 | |
| *** josecastroleon has joined #openstack-nova | 04:53 | |
| *** josecastroleon has quit IRC | 04:55 | |
| *** nkrinner has joined #openstack-nova | 04:58 | |
| *** markvoelker has quit IRC | 04:58 | |
| *** ildikov has joined #openstack-nova | 04:59 | |
| *** CongTo has joined #openstack-nova | 04:59 | |
| *** _gryf|afk is now known as _gryf | 05:01 | |
| *** CongToThanh has joined #openstack-nova | 05:02 | |
| *** zhangtralon has joined #openstack-nova | 05:02 | |
| *** CongToThanh has quit IRC | 05:02 | |
| *** patrickeast has quit IRC | 05:03 | |
| *** markvoelker has joined #openstack-nova | 05:06 | |
| *** ir2ivps3 has quit IRC | 05:06 | |
| *** unicell has joined #openstack-nova | 05:07 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Raise invalid input if use invalid ip for network to attach interface https://review.openstack.org/162894 | 05:07 |
| *** rook has quit IRC | 05:08 | |
| *** zhangtralon has quit IRC | 05:08 | |
| *** unicell has quit IRC | 05:11 | |
| *** vilobhmm11 has joined #openstack-nova | 05:11 | |
| *** unicell has joined #openstack-nova | 05:12 | |
| *** achanda_ has joined #openstack-nova | 05:12 | |
| *** hemna_ has joined #openstack-nova | 05:13 | |
| *** vilobhmm1 has quit IRC | 05:15 | |
| *** rook has joined #openstack-nova | 05:15 | |
| *** yingjun has quit IRC | 05:16 | |
| *** mikal_ has joined #openstack-nova | 05:17 | |
| *** Sree_ has joined #openstack-nova | 05:18 | |
| *** dosaboy_ has joined #openstack-nova | 05:18 | |
| *** thomasem_ has joined #openstack-nova | 05:19 | |
| *** Sree_ is now known as Guest23384 | 05:19 | |
| *** mishkin___ has joined #openstack-nova | 05:19 | |
| *** redbeard has joined #openstack-nova | 05:20 | |
| *** coolsvap- has joined #openstack-nova | 05:23 | |
| *** ir2ivps3 has joined #openstack-nova | 05:24 | |
| *** bkopilov is now known as bkopilov_wfh | 05:24 | |
| *** fawadkhaliq has joined #openstack-nova | 05:24 | |
| *** eliqiao1 has joined #openstack-nova | 05:25 | |
| *** achanda has quit IRC | 05:26 | |
| *** coolsvap has quit IRC | 05:26 | |
| *** mishkin__ has quit IRC | 05:26 | |
| *** hemna has quit IRC | 05:26 | |
| *** anteaya has quit IRC | 05:26 | |
| *** thomasem has quit IRC | 05:26 | |
| *** eliqiao has quit IRC | 05:26 | |
| *** dosaboy has quit IRC | 05:27 | |
| *** mikal has quit IRC | 05:27 | |
| openstackgerrit | Sudipta Biswas proposed openstack/nova-specs: Cleanup 'scheduled_at' from instances table https://review.openstack.org/180120 | 05:27 |
| *** anteaya has joined #openstack-nova | 05:29 | |
| *** CongTo has quit IRC | 05:29 | |
| *** herch has joined #openstack-nova | 05:30 | |
| *** CongTo has joined #openstack-nova | 05:30 | |
| *** Guest23384 has quit IRC | 05:32 | |
| *** rook has quit IRC | 05:34 | |
| *** Nikolay_St has quit IRC | 05:34 | |
| herch | hi, anyway to have multiple hypervisors on same host and still be able to manage their VMs using openstack? | 05:36 |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge test_admin_actions between v2 and v2.1 https://review.openstack.org/170754 | 05:39 |
| openstackgerrit | IWAMOTO Toshihiro proposed openstack/nova: Omit RescheduledException in instance_fault.message https://review.openstack.org/165751 | 05:45 |
| *** kiranr has joined #openstack-nova | 05:45 | |
| *** claudiub has joined #openstack-nova | 05:45 | |
| *** ildikov has quit IRC | 05:45 | |
| *** dulek has joined #openstack-nova | 05:48 | |
| *** josecastroleon has joined #openstack-nova | 05:50 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge V2 and V2.1 hypervisor functional tests https://review.openstack.org/167082 | 05:50 |
| *** claudiub has quit IRC | 05:51 | |
| *** lpetrut has joined #openstack-nova | 05:56 | |
| *** dims has joined #openstack-nova | 06:01 | |
| *** yingjun has joined #openstack-nova | 06:02 | |
| *** kiranr is now known as kiran-r | 06:04 | |
| *** CongTo has quit IRC | 06:06 | |
| *** dims has quit IRC | 06:06 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Refactor show_port() in neutron api https://review.openstack.org/135020 | 06:09 |
| *** afazekas_ has joined #openstack-nova | 06:11 | |
| *** park_ has joined #openstack-nova | 06:16 | |
| *** sahid has joined #openstack-nova | 06:17 | |
| openstackgerrit | Marian Horban proposed openstack/nova: Disassociating of floating IPs corrected https://review.openstack.org/149358 | 06:22 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Imported Translations from Transifex https://review.openstack.org/179638 | 06:23 |
| *** mpaolino has joined #openstack-nova | 06:23 | |
| *** otter768 has joined #openstack-nova | 06:23 | |
| openstackgerrit | melanie witt proposed openstack/nova: Send Instance object to cells instance_update_at_top https://review.openstack.org/176518 | 06:25 |
| *** jcoufal has joined #openstack-nova | 06:25 | |
| *** deepthi has quit IRC | 06:26 | |
| *** otter768 has quit IRC | 06:28 | |
| *** moshele has joined #openstack-nova | 06:31 | |
| *** CongTo has joined #openstack-nova | 06:31 | |
| *** ildikov has joined #openstack-nova | 06:32 | |
| *** VW_ has quit IRC | 06:33 | |
| openstackgerrit | Eli Qiao proposed openstack/nova: Object: Merge db accesses for extra field of instance https://review.openstack.org/160076 | 06:34 |
| *** Murali has joined #openstack-nova | 06:37 | |
| *** deepthi has joined #openstack-nova | 06:37 | |
| *** asettle has quit IRC | 06:38 | |
| *** VW has joined #openstack-nova | 06:45 | |
| *** zhangtralon has joined #openstack-nova | 06:49 | |
| *** lpetrut has quit IRC | 06:50 | |
| *** irenab has joined #openstack-nova | 06:51 | |
| *** pece has joined #openstack-nova | 06:51 | |
| *** dulek has quit IRC | 06:51 | |
| *** dulek has joined #openstack-nova | 06:51 | |
| *** kiranr has joined #openstack-nova | 06:54 | |
| *** achanda_ has quit IRC | 06:55 | |
| *** fifieldt has joined #openstack-nova | 06:56 | |
| *** fifieldt has quit IRC | 06:56 | |
| openstackgerrit | Oleg Bondarev proposed openstack/nova: Log a warning if sec-groups and port id are provided on boot https://review.openstack.org/154068 | 06:56 |
| *** kiran-r has quit IRC | 06:57 | |
| *** herch has quit IRC | 06:58 | |
| openstackgerrit | Roman Bogorodskiy proposed openstack/nova-specs: NIC bonding support for SRIOV ports https://review.openstack.org/182242 | 06:59 |
| openstackgerrit | melanie witt proposed openstack/nova: Send Instance object to cells instance_update_at_top https://review.openstack.org/176518 | 07:01 |
| *** cfriesen has quit IRC | 07:02 | |
| *** kiranr has quit IRC | 07:02 | |
| *** kiranr has joined #openstack-nova | 07:02 | |
| openstackgerrit | Balazs Gibizer proposed openstack/nova-specs: Add notification for service status change https://review.openstack.org/182350 | 07:03 |
| *** markus_z has joined #openstack-nova | 07:06 | |
| *** CongTo has quit IRC | 07:07 | |
| openstackgerrit | Rui Chen proposed openstack/nova: Fix resource leaking when consume_from_instance raise exception https://review.openstack.org/182165 | 07:07 |
| *** ZZelle_ has quit IRC | 07:08 | |
| *** ZZelle_ has joined #openstack-nova | 07:08 | |
| *** dulek has quit IRC | 07:09 | |
| *** ildikov has quit IRC | 07:10 | |
| *** VW has quit IRC | 07:13 | |
| *** jichen has joined #openstack-nova | 07:16 | |
| *** kiranr is now known as kiran-r | 07:16 | |
| *** doron_afk has joined #openstack-nova | 07:18 | |
| *** mpavone has joined #openstack-nova | 07:18 | |
| *** x3k is now known as xek | 07:19 | |
| *** redbeard has quit IRC | 07:22 | |
| *** ildikov has joined #openstack-nova | 07:25 | |
| *** eglynn-afk has quit IRC | 07:25 | |
| *** vilobhmm11 has quit IRC | 07:27 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Make resize api of compute manager to send object https://review.openstack.org/161860 | 07:29 |
| *** kodoku has joined #openstack-nova | 07:30 | |
| *** Marga_ has quit IRC | 07:31 | |
| *** Nikolay_St has joined #openstack-nova | 07:31 | |
| *** jlanoux has joined #openstack-nova | 07:38 | |
| *** deepthi has quit IRC | 07:39 | |
| *** xek has quit IRC | 07:40 | |
| *** xek has joined #openstack-nova | 07:41 | |
| *** deepthi has joined #openstack-nova | 07:44 | |
| *** gszasz has joined #openstack-nova | 07:46 | |
| *** jistr has joined #openstack-nova | 07:49 | |
| *** zhangtralon has quit IRC | 07:49 | |
| *** zhangtralon has joined #openstack-nova | 07:50 | |
| *** dulek has joined #openstack-nova | 07:50 | |
| *** danpb has joined #openstack-nova | 07:51 | |
| *** browne has quit IRC | 07:54 | |
| *** dulek has quit IRC | 07:55 | |
| *** mmedvede has quit IRC | 07:56 | |
| *** mpavone has quit IRC | 07:57 | |
| *** dsanders has joined #openstack-nova | 07:58 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Make resize api of compute manager to send object https://review.openstack.org/161860 | 07:59 |
| *** Longgeek has joined #openstack-nova | 08:00 | |
| *** eglynn-afk has joined #openstack-nova | 08:01 | |
| *** chaochin has joined #openstack-nova | 08:03 | |
| *** derekh has joined #openstack-nova | 08:03 | |
| *** mpavone has joined #openstack-nova | 08:04 | |
| *** mmedvede has joined #openstack-nova | 08:11 | |
| johnthetubaguy | dansmith: jogo: thanks for pushing that through for me, a hopped off after the cross project meeting | 08:12 |
| openstackgerrit | Tina Tang proposed openstack/nova: Fix in LibvirtISCSIVolumeDriver to avoid deleting incorrect device https://review.openstack.org/182565 | 08:12 |
| johnthetubaguy | deepakcs: I am around now (although just running downstairs for a coffee), can I help? | 08:13 |
| *** ppal has joined #openstack-nova | 08:14 | |
| deepakcs | johnthetubaguy: reg. your comment in https://review.openstack.org/#/c/134517/7/specs/liberty/approved/volume-snapshot-improvements.rst | 08:15 |
| deepakcs | johnthetubaguy: i commented back. I don't think that spec is a dep, its related tho' | 08:16 |
| johnthetubaguy | deepakcs: OK, thats cool, thanks | 08:16 |
| *** dsanders has quit IRC | 08:18 | |
| bauzas | johnthetubaguy: I'm just reading the current WIP devrefs, happy with comments if they are -W ? | 08:18 |
| bauzas | even if, should I say rather | 08:18 |
| *** yingjun has quit IRC | 08:18 | |
| bauzas | flavors and errors are the ones I'm digging into | 08:18 |
| johnthetubaguy | bauzas: all comments welcome, just got the first one merged, so I was going to go back and look at those later on | 08:19 |
| johnthetubaguy | bauzas: thats cool | 08:19 |
| bauzas | johnthetubaguy: cool, will do | 08:19 |
| *** eglynn-afk is now known as eglynn | 08:19 | |
| bauzas | johnthetubaguy: do you know the philly ops etherpad which talked about errors ? | 08:19 |
| *** haypo has joined #openstack-nova | 08:19 | |
| bauzas | johnthetubaguy: https://etherpad.openstack.org/p/PHL-ops-nova-feedback | 08:20 |
| johnthetubaguy | bauzas: I read it one upon a time, about the "self debugging" stuff, thats partly what the summit session hopes to cover | 08:20 |
| johnthetubaguy | bauzas: just nipping to get a coffee, brb | 08:20 |
| bauzas | johnthetubaguy: okay, I thought I would probably add some comments on the errors devref change to make it clearer | 08:20 |
| bauzas | johnthetubaguy: sure thing, you had short night | 08:20 |
| kashyap | It'd be nice the plain text exported versions of these are posted somewhere for permanancy. Etherpads are very sluggish. | 08:21 |
| kashyap | s/the/if the | 08:21 |
| bauzas | kashyap: dansmith told me once how to make those plain text, sec | 08:21 |
| *** oro_ has quit IRC | 08:21 | |
| kashyap | Just export it. | 08:21 |
| *** oro has quit IRC | 08:21 | |
| bauzas | kashyap: yeah https://etherpad.openstack.org/p/PHL-ops-nova-feedback/export/txt | 08:23 |
| kashyap | Yep. The point being, over the period of a cycle plenty of these will be accumulated, searching Etherpads for that one bit of info will become cumbersome. . . | 08:23 |
| *** otter768 has joined #openstack-nova | 08:24 | |
| sahid | danpb: about the work on libvirt, i'm working on making Host.get_domain to return a Guest object, is that makes sense for you or do you think it will be better to keep sepration between Host and Guest? | 08:24 |
| *** nihilifer has joined #openstack-nova | 08:25 | |
| *** otter768 has quit IRC | 08:29 | |
| *** lucasagomes has joined #openstack-nova | 08:29 | |
| *** kodoku has quit IRC | 08:30 | |
| johnthetubaguy | bauzas: comments are good, I wrote that on an ipad on a train, so its not exactly ready yet, but comments would be good | 08:31 |
| bauzas | johnthetubaguy: yeah, my comments are really on how we structure those docs | 08:32 |
| bauzas | johnthetubaguy: IMHO that should be both a reference document and guidelines for future changes | 08:32 |
| bauzas | johnthetubaguy: like a manifesto for improvement | 08:32 |
| johnthetubaguy | bauzas: thats the plan, it doesn't have the second bit written yet | 08:33 |
| bauzas | johnthetubaguy: not exactly as detailed as a spec, but at least covering the limitations and stating the possible efforts | 08:33 |
| johnthetubaguy | bauzas: I was hoping to have the "current" state merged before the summit, but well, thats not happening | 08:33 |
| bauzas | johnthetubaguy: ok, then I just braindumped a few things | 08:33 |
| *** oro_ has joined #openstack-nova | 08:33 | |
| *** oro has joined #openstack-nova | 08:33 | |
| bauzas | johnthetubaguy: IMHO those docs are good for providing contributors a way to improve Nova on a consensual thing | 08:34 |
| johnthetubaguy | bauzas: +1 hoping to do loads more of these | 08:35 |
| bauzas | johnthetubaguy: yeah I like the idea | 08:35 |
| *** Jeffrey4l has quit IRC | 08:36 | |
| bauzas | johnthetubaguy: and you have all my credits for having proposed a change to the architectural diagram | 08:36 |
| bauzas | johnthetubaguy: for that, I mean, it was just *necessary* | 08:36 |
| johnthetubaguy | cools | 08:37 |
| openstackgerrit | Rui Chen proposed openstack/python-novaclient: server-group-list support 'all_projects' parameter https://review.openstack.org/182568 | 08:38 |
| openstackgerrit | Nisha Agarwal proposed openstack/nova-specs: Nova and Ironic-virt-driver to support json list https://review.openstack.org/182572 | 08:42 |
| *** dguitarbite has joined #openstack-nova | 08:43 | |
| openstackgerrit | Lan Qi Song proposed openstack/nova: Remove useless volume when boot from volume failed https://review.openstack.org/177084 | 08:46 |
| *** Jeffrey4l has joined #openstack-nova | 08:47 | |
| openstackgerrit | sahid proposed openstack/nova: libvirt: make _create_domain to return a Guest object https://review.openstack.org/182185 | 08:49 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method shutdown to Guest object https://review.openstack.org/182194 | 08:49 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method resume to Guest object https://review.openstack.org/182195 | 08:49 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce a Guest to wrap around virConnect https://review.openstack.org/174313 | 08:49 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method get_guest to Host object https://review.openstack.org/181969 | 08:49 |
| *** josecastroleon has quit IRC | 08:50 | |
| sahid | johnthetubaguy: hello, any chance you take a look at my coments on the spec here https://review.openstack.org/#/c/165838/ ? | 08:51 |
| *** sudipto has joined #openstack-nova | 08:51 | |
| *** yingjun has joined #openstack-nova | 08:51 | |
| *** ericksonfgds has joined #openstack-nova | 08:52 | |
| johnthetubaguy | sahid: so the token as a hash thing | 08:53 |
| gilliard | alex_xu: are you there? | 08:53 |
| johnthetubaguy | sahid: I was thinking, if you have the token in the DB, and you get access to the DB, you could use the token | 08:53 |
| johnthetubaguy | sahid: if you get access to only the hash, its no use to the attaker | 08:53 |
| *** josecastroleon has joined #openstack-nova | 08:54 | |
| danpb | sahid: yes, I would like Host to return Guest instances, instead of returning libvirt.virDomain instances | 08:55 |
| deepakcs | johnthetubaguy: thanks for +2ing the spec patch | 08:55 |
| danpb | sahid: as that matches our general goal that we isolate the places which talk to the libvirt python module in just host.py and guest.py | 08:55 |
| sahid | johnthetubaguy: ok so the token will be in "clear" on when the user call the API to retrieve it.. I understand now | 08:55 |
| danpb | the rest of the nova code should exclusively use the Host and Guest classes and not the libvirt python classes | 08:56 |
| johnthetubaguy | sahid: maybe I don't get how that all works, but I was more thinking about people looking at the tokens on the disk | 08:56 |
| sahid | johnthetubaguy: well mysqk store data in binary i do not believe it easy to read token on the disk | 08:57 |
| sahid | i mean it's probably more dificult than read them from memcached | 08:57 |
| johnthetubaguy | sahid: did you see some else also had suggested storing this in the DB | 08:58 |
| sahid | johnthetubaguy: you mena the other spec you have pointed ? | 08:58 |
| johnthetubaguy | sahid: its on the disk, you can read that in many many ways, memcache you need to get access to the memory while it contains the tokens, I would thinking the exact opposite | 08:58 |
| sahid | mena/mean | 08:58 |
| johnthetubaguy | sahid: yep | 08:59 |
| sahid | johnthetubaguy: memcched do not provide any security layer,you can just use the api on the host | 08:59 |
| sahid | we ctually use the instance_uuid as key | 08:59 |
| johnthetubaguy | sahid: right, thats a deployment problem, easily fixed | 08:59 |
| sahid | johnthetubaguy: but yes you solution about using a hash i interesting | 09:00 |
| sahid | your | 09:00 |
| sahid | is | 09:00 |
| sahid | i will refer it on the spec | 09:00 |
| johnthetubaguy | sahid: so if we need to give back users tokens we already have in the DB, then a hash would not work, unless each request gets a new token (which honestly, is my preference) | 09:00 |
| johnthetubaguy | sahid: the problem is storing the token on disk, rather than it just living in memory, thats quite a change, security wise | 09:01 |
| sahid | johnthetubaguy: ok by you solution we will improve the security and it is something important we can do | 09:01 |
| sahid | sure | 09:01 |
| johnthetubaguy | sahid: well, we should get a security team expert to review this | 09:01 |
| johnthetubaguy | sahid: probably need to add a salt to the hash, etc | 09:01 |
| sahid | yes.. | 09:02 |
| sahid | let me rework the spec to add your thinking and we will see | 09:02 |
| *** doron_afk is now known as doron | 09:03 | |
| sahid | danpb: cool that is what i was thinking but i wanted to get your advice | 09:04 |
| sahid | danpb: i have started this work here https://review.openstack.org/#/c/181969/ | 09:04 |
| danpb | yes, just saw that | 09:06 |
| johnthetubaguy | sahid: FWIW, for the folks that like using memcache, I worry about us just dropping that support here, and using the DB instead, but I like the idea of the DB as the new default | 09:06 |
| *** ajo has joined #openstack-nova | 09:07 | |
| ajo | bauzas ping | 09:07 |
| ajo | quick question | 09:07 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova: scheduler:refactor group info in filter_properties https://review.openstack.org/148275 | 09:07 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova: api: add soft-affinity policies for server groups https://review.openstack.org/147520 | 09:07 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova: Add functional tests to soft affinity capabilities https://review.openstack.org/147521 | 09:07 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova: scheduler: add num of instance per host in a group https://review.openstack.org/148277 | 09:07 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova: object:count members per hosts in InstanceGroup https://review.openstack.org/147517 | 09:07 |
| ajo | is there a way to force nova to start an instance in a certain compute node? | 09:07 |
| bauzas | ajo | 09:07 |
| bauzas | ajo: yup | 09:08 |
| sahid | johnthetubaguy: yes, i'm thinking - if we want to keep working both that mean maintenant the old code + add support for the framework objects and make them working together | 09:08 |
| kashyap | ajo: I was testing something like that yesterday: | 09:08 |
| bauzas | ajo: --availability-zone=nova:host | 09:08 |
| *** Jeffrey4l has quit IRC | 09:08 | |
| kashyap | $ nova boot --config-drive false --flavor 1 --key_name oskey1 --image cirros-0.3.3-x86_64-disk --availability-zone=nova:devstack3 cirrvm11-as-demo | 09:08 |
| ajo | bauzas++ | 09:08 |
| ajo | thanks, I was in a run | 09:08 |
| ajo | thanks kashyap :) | 09:08 |
| ajo | you're awesome guys :) | 09:09 |
| kashyap | ajo: Note that it only works as an admin role. | 09:09 |
| *** dulek has joined #openstack-nova | 09:09 | |
| johnthetubaguy | sahid: we can refactor the old memcache code to like like a DB API, then just switch in and out the DB API used by the objects right? | 09:09 |
| johnthetubaguy | sahid: at least thats what I was thinking about when I was suggesting we support both | 09:09 |
| openstackgerrit | Nisha Agarwal proposed openstack/nova-specs: Add IronicFilter for ironic act-upon capabilities https://review.openstack.org/182581 | 09:10 |
| *** Jeffrey4l has joined #openstack-nova | 09:10 | |
| johnthetubaguy | sahid: now we could deprecate the memcache support when you add the DB support, I am OK with that, its the removing it without deprecating (and a chance for feedback) that worries me | 09:12 |
| sahid | johnthetubaguy: well yes problably i do not know if "just switch un and out DB API" is the good term here but yes it's probably the way to follow :) | 09:13 |
| johnthetubaguy | sahid: :) "persistence layer" maybe, but yeah | 09:13 |
| sahid | johnthetubaguy: please comment on the review your thinking on the hash point and memcache support like that i will take the time to investigate how we can do that | 09:15 |
| sahid | to keep a trace | 09:15 |
| johnthetubaguy | sahid: yeah, I added comments while we were talking here, let me know if I missed something | 09:16 |
| sahid | sure thank you | 09:17 |
| johnthetubaguy | sahid: the tokens have an expiry time, which helps reduce the impact of not doing the hash, but it would be good to talk about rejecting the idea in the alternatives if its not practical | 09:17 |
| openstackgerrit | Victor Stinner proposed openstack/nova: Replace iter.next() with next(iter) https://review.openstack.org/179821 | 09:18 |
| sahid | johnthetubaguy: well from how i see that we can probably do it, but yes we need to investigate to see if no blockers | 09:18 |
| *** ankit_ag has joined #openstack-nova | 09:19 | |
| sudipto | bauzas, Hi, i was looking at scrubbing through some of the old Nova defects and found this. https://bugs.launchpad.net/nova/+bug/1369818 I actually think this is fixed in some form in Kilo. Your thoughts? | 09:20 |
| openstack | Launchpad bug 1369818 in OpenStack Compute (nova) "Nova show will not display NoValidHost with right exception traces" [Medium,Confirmed] - Assigned to Sudipta Biswas (sbiswas7) | 09:20 |
| johnthetubaguy | sahid: cool, I am guessing it might mean we generate a new token for every request for a token, while wasteful, I quite like it | 09:20 |
| *** ericksonfgds has quit IRC | 09:20 | |
| openstackgerrit | Victor Stinner proposed openstack/nova: Fix raise syntax for Python 3 https://review.openstack.org/179946 | 09:27 |
| *** dulek has quit IRC | 09:28 | |
| *** dulek has joined #openstack-nova | 09:29 | |
| Anticimex | does nova-api, nova-compute and nova-conductor's MQ client libs handle multiple MQ endpoints? | 09:32 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method poweroff to Guest object https://review.openstack.org/182194 | 09:35 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method resume to Guest object https://review.openstack.org/182195 | 09:35 |
| *** dims has joined #openstack-nova | 09:35 | |
| *** fawadkhaliq has quit IRC | 09:36 | |
| *** yingjun has quit IRC | 09:37 | |
| *** doron is now known as doron_afk | 09:39 | |
| *** apuimedo has joined #openstack-nova | 09:40 | |
| *** 5EXAA60VI has quit IRC | 09:40 | |
| *** dims has quit IRC | 09:40 | |
| openstackgerrit | Victor Stinner proposed openstack/nova: Avoid useless copy in get_instance_metadata() https://review.openstack.org/182596 | 09:40 |
| *** gszasz has quit IRC | 09:41 | |
| *** ildikov has quit IRC | 09:48 | |
| johnthetubaguy | Anticimex: sorry, not sure I understand your question, do you mean having multiple queuing servers? | 09:52 |
| Anticimex | yes | 09:53 |
| Anticimex | e.g. pointing to multiple endpoints of the same MQ cluster | 09:53 |
| *** CongTo has joined #openstack-nova | 09:53 | |
| openstackgerrit | Michal Rostecki proposed openstack/nova: Create a reserved aggregate filter https://review.openstack.org/178750 | 09:54 |
| *** tojuvone has quit IRC | 09:55 | |
| openstackgerrit | Nisha Agarwal proposed openstack/nova-specs: Add IronicFilter for ironic act-upon capabilities https://review.openstack.org/182581 | 09:56 |
| openstackgerrit | Nisha Agarwal proposed openstack/nova-specs: Nova and Ironic-virt-driver to support json list https://review.openstack.org/182572 | 09:59 |
| *** moshele has quit IRC | 10:00 | |
| *** tojuvone has joined #openstack-nova | 10:00 | |
| *** moshele has joined #openstack-nova | 10:01 | |
| *** dims has joined #openstack-nova | 10:01 | |
| *** lpetrut has joined #openstack-nova | 10:02 | |
| *** ildikov has joined #openstack-nova | 10:03 | |
| *** IanGovett has joined #openstack-nova | 10:06 | |
| sudipto | johnthetubaguy, bauzas I actually found somewhat of a miss match while investigating the bug 1369818...It appears that there's a bit of an inconsistency in the way we show the "NoValidHost" exception to the user. in the scheduler/filter_scheduler.py - while raising the exception - we are saying we shouldn't include much information about the environment to the user. But at the same time, in the nova/scheduler/utils.py - if the max_attempts have e | 10:07 |
| openstack | bug 1369818 in OpenStack Compute (nova) "Nova show will not display NoValidHost with right exception traces" [Medium,Confirmed] https://launchpad.net/bugs/1369818 - Assigned to Sudipta Biswas (sbiswas7) | 10:07 |
| sudipto | xceeded - we are showing the full stack details of what went wrong with the last host - as a part of the exception... | 10:07 |
| johnthetubaguy | sudipto: yeah, we shouldn't be giving the user a full stack trace | 10:07 |
| johnthetubaguy | sudipto: but we want that in notifications and instance_faults, etc | 10:08 |
| *** jamielennox is now known as jamielennox|away | 10:08 | |
| johnthetubaguy | sudipto: its a mess around that error, been meaning to take a closer look for a little while now | 10:08 |
| bauzas | johnthetubaguy: sudipto: there is another bug that mentions that | 10:08 |
| bauzas | sudipto: https://bugs.launchpad.net/nova/+bug/1431291 | 10:09 |
| openstack | Launchpad bug 1431291 in OpenStack Compute (nova) "Scheduler Failures are no longer logged with enough detail for a site admin to do problem determination" [High,In progress] - Assigned to Sylvain Bauza (sylvain-bauza) | 10:09 |
| sudipto | johnthetubaguy, bauzas ok - https://github.com/openstack/nova/blob/master/nova/scheduler/utils.py#L165 | 10:09 |
| bauzas | sudipto: the main problem is how the scheduler can know the instance.uuid he is working on | 10:09 |
| *** doude has joined #openstack-nova | 10:09 | |
| bauzas | sudipto: IMHO you're working on a bug that needs further discussion | 10:10 |
| sudipto | bauzas, Yup - so i pinged you earlier :D | 10:10 |
| bauzas | sudipto: we should be discussing on how to improve that feedback during the Summit | 10:10 |
| sudipto | bauzas, :( won't be at the summit but i will probably talk to you on IRC | 10:11 |
| bauzas | sudipto: yeah, I have ideas and POCs, but I need to share that with the team | 10:11 |
| bauzas | sudipto: what johnthetubaguy explained is one of the things that I would see improvements | 10:11 |
| bauzas | sudipto: ie. instance faults | 10:11 |
| bauzas | sudipto: the story is like this, the scheduler doesn't know which instances he has to decide | 10:12 |
| bauzas | sudipto: so, he runs thru all the filters and give back to the conductor a list of hosts or None | 10:12 |
| johnthetubaguy | bauzas: I think thats the wrong way around, the conductor needs to do the error handling for the error the scheduler returns, but its a little mixed up right now I suspect | 10:13 |
| bauzas | johnthetubaguy: that's exactly what I'm writing :) | 10:13 |
| johnthetubaguy | bauzas: I think you are mixing up returning zero hosts, and retrying three times due to exceptions | 10:13 |
| bauzas | johnthetubaguy: but the conductor doesn't know what happened in the scheduler | 10:13 |
| johnthetubaguy | agreed | 10:13 |
| bauzas | johnthetubaguy: that's another story | 10:13 |
| johnthetubaguy | bauzas: agreed, but its the same error right now, which is a big issue, having those errors different would be a good step forward by its-self | 10:14 |
| bauzas | johnthetubaguy: having a same NoValidHost exception is confusing | 10:14 |
| bauzas | eh, jinxed | 10:14 |
| johnthetubaguy | :) | 10:14 |
| *** gszasz has joined #openstack-nova | 10:15 | |
| johnthetubaguy | bauzas: turns out they look slightly different right now, but they should be totally different | 10:15 |
| bauzas | agreed | 10:15 |
| sudipto | there's one more case I see (and excuse my little knowledge) - the conductor/manager.py - has the build_instances method. I have a total of 2 hosts and my max_retries is set to > 2. In such a case, even though we exhaust the 2 hosts to try on and fail in conductor - we still end up calling the scheduler_client.select_destinations method. | 10:15 |
| sudipto | it probably is just an optimization - but still... | 10:15 |
| bauzas | sudipto: compute.build_instances() will burn in hell soon | 10:16 |
| sudipto | bauzas, Ah i see.. | 10:16 |
| sudipto | bauzas, So for the time being, I guess i should just hold on to the bug i have and wait for your discussion at the summit for a way forward. | 10:16 |
| bauzas | sudipto: well, the problem is that it's not trivial to fix :/ | 10:17 |
| bauzas | I mean, fixing that correctly | 10:17 |
| sudipto | bauzas, yeah i just started debugging a few bugs based on an area interest per johnthetubaguy 's suggestion and ended up in this place, where i quite frankly had to struggle a bit to get the hang of what was going on here. | 10:17 |
| bauzas | sudipto: that's all your honor, sir. | 10:18 |
| sudipto | bauzas, I'd be interested if you guys come to some conclusion during the summit and need some volunteers. I wouldn't be there sigh! | 10:18 |
| bauzas | sudipto: sure thing | 10:18 |
| *** dulek has quit IRC | 10:19 | |
| bauzas | johnthetubaguy: oh I see why you thought I made a confusion, since I was explaining the single scheduler failure story | 10:20 |
| sudipto | johnthetubaguy, bauzas on a separate area, why do we have these auth urls in the nova.conf - instead of retrieving them from the service catalog? | 10:23 |
| *** wanghao has quit IRC | 10:25 | |
| *** melwitt has quit IRC | 10:25 | |
| *** josecastroleon1 has joined #openstack-nova | 10:25 | |
| *** otter768 has joined #openstack-nova | 10:25 | |
| openstackgerrit | jichenjc proposed openstack/nova: Change group controller of V2 test cases https://review.openstack.org/182613 | 10:25 |
| openstackgerrit | jichenjc proposed openstack/nova: Add schema check for security_groups https://review.openstack.org/182614 | 10:25 |
| *** wanghao has joined #openstack-nova | 10:25 | |
| *** jhesketh has quit IRC | 10:25 | |
| *** wendar has quit IRC | 10:25 | |
| *** gus has quit IRC | 10:25 | |
| johnthetubaguy | sudipto: sometimes you want to use a different one to the user catalog, mostly just because that was easy at the time. Defaulting to the service catalog might make a lot of sense as a spec | 10:25 |
| *** wendar_ has joined #openstack-nova | 10:25 | |
| *** GLaupre has quit IRC | 10:26 | |
| *** gszasz has quit IRC | 10:26 | |
| *** jbernard has quit IRC | 10:26 | |
| *** obondarev_ has quit IRC | 10:26 | |
| sudipto | johnthetubaguy, ok... | 10:27 |
| *** tlbr has quit IRC | 10:27 | |
| sudipto | johnthetubaguy, so it's mostly a evolution problem. | 10:27 |
| *** obondarev has joined #openstack-nova | 10:27 | |
| *** melwitt has joined #openstack-nova | 10:28 | |
| *** mikal_ has quit IRC | 10:28 | |
| *** mikal has joined #openstack-nova | 10:28 | |
| *** josecastroleon has quit IRC | 10:28 | |
| johnthetubaguy | sudipto: its mostly something no one has cared enough about to fix / think through | 10:28 |
| *** Vek has quit IRC | 10:28 | |
| *** dulek has joined #openstack-nova | 10:28 | |
| sudipto | johnthetubaguy, ok.. | 10:29 |
| *** josecastroleon1 has quit IRC | 10:29 | |
| *** xgsa has quit IRC | 10:30 | |
| *** promulo_ has quit IRC | 10:30 | |
| *** otter768 has quit IRC | 10:30 | |
| *** promulo_ has joined #openstack-nova | 10:30 | |
| *** xgsa has joined #openstack-nova | 10:30 | |
| *** jhesketh has joined #openstack-nova | 10:31 | |
| johnthetubaguy | sudipto: would be happy to see that config sorted out though | 10:31 |
| johnthetubaguy | sudipto: really looking for someone to go through and tag all the config variables with a classification, or something similar | 10:32 |
| *** EmilienM has quit IRC | 10:33 | |
| *** danieru has quit IRC | 10:33 | |
| sudipto | johnthetubaguy, ok i can give it a shot. | 10:33 |
| *** tlbr has joined #openstack-nova | 10:33 | |
| *** ekarlso has quit IRC | 10:33 | |
| *** gus has joined #openstack-nova | 10:33 | |
| *** pixelbeat_ has joined #openstack-nova | 10:33 | |
| *** danieru has joined #openstack-nova | 10:34 | |
| *** EmilienM has joined #openstack-nova | 10:35 | |
| *** jbernard has joined #openstack-nova | 10:36 | |
| *** deepthi has quit IRC | 10:37 | |
| openstackgerrit | sahid proposed openstack/nova: libvirt: make _create_domain to return a Guest object https://review.openstack.org/182185 | 10:40 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method poweroff to Guest object https://review.openstack.org/182194 | 10:40 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method resume to Guest object https://review.openstack.org/182195 | 10:40 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: introduce method get_guest to Host object https://review.openstack.org/181969 | 10:40 |
| *** ekarlso has joined #openstack-nova | 10:40 | |
| *** cbader has quit IRC | 10:41 | |
| *** deepakcs has quit IRC | 10:41 | |
| *** mpaolino has quit IRC | 10:44 | |
| *** ekarlso has quit IRC | 10:45 | |
| *** ekarlso has joined #openstack-nova | 10:45 | |
| openstackgerrit | Stuart McLaren proposed openstack/nova: Add explicit dependency on python-keystoneclient https://review.openstack.org/181966 | 10:47 |
| johnthetubaguy | bauzas: what do you mean by "context request id is not persisted"? | 10:49 |
| *** zhangtralon has quit IRC | 10:49 | |
| *** dulek has quit IRC | 10:50 | |
| *** dulek has joined #openstack-nova | 10:50 | |
| openstackgerrit | John Garbutt proposed openstack/nova: devref: discuss current state of error handling https://review.openstack.org/178635 | 10:51 |
| *** dulek_ has joined #openstack-nova | 10:53 | |
| *** dulek_ has quit IRC | 10:53 | |
| *** dulek_ has joined #openstack-nova | 10:54 | |
| *** dulek has quit IRC | 10:54 | |
| *** kmartin has quit IRC | 10:56 | |
| *** lucasagomes is now known as lucas-afk | 10:57 | |
| ekarlso | johnthetubaguy: heya | 11:00 |
| ekarlso | johnthetubaguy: you up for a nova q ? | 11:00 |
| *** asselin has quit IRC | 11:01 | |
| *** ildikov has quit IRC | 11:02 | |
| johnthetubaguy | ekarlso: sure | 11:06 |
| *** moshele has quit IRC | 11:07 | |
| openstackgerrit | Mikhail Feoktistov proposed openstack/nova: Virtuozzo instance resize support https://review.openstack.org/182257 | 11:11 |
| *** bfic has joined #openstack-nova | 11:12 | |
| *** ociuhandu has joined #openstack-nova | 11:13 | |
| *** ildikov has joined #openstack-nova | 11:17 | |
| *** Jeffrey4l has quit IRC | 11:22 | |
| *** salv-orlando has joined #openstack-nova | 11:28 | |
| *** moshele has joined #openstack-nova | 11:29 | |
| *** sahid has quit IRC | 11:31 | |
| *** VW has joined #openstack-nova | 11:31 | |
| *** VW has quit IRC | 11:32 | |
| *** VW has joined #openstack-nova | 11:32 | |
| *** sahid has joined #openstack-nova | 11:34 | |
| *** mjturek1 has joined #openstack-nova | 11:34 | |
| *** tobe4333 has joined #openstack-nova | 11:35 | |
| *** atuvenie has joined #openstack-nova | 11:36 | |
| *** josecastroleon has joined #openstack-nova | 11:36 | |
| *** salv-orlando has quit IRC | 11:37 | |
| *** moshele has quit IRC | 11:38 | |
| *** Jeffrey4l has joined #openstack-nova | 11:38 | |
| *** moshele has joined #openstack-nova | 11:39 | |
| openstackgerrit | Mikhail Feoktistov proposed openstack/nova: Virtuozzo instance resize support https://review.openstack.org/182257 | 11:40 |
| *** coolsvap- is now known as coolsvap|afk | 11:41 | |
| *** salv-orlando has joined #openstack-nova | 11:43 | |
| *** IanGovett has quit IRC | 11:43 | |
| *** artom has joined #openstack-nova | 11:46 | |
| *** mwagner_lap has quit IRC | 11:46 | |
| *** yamahata has quit IRC | 11:48 | |
| *** IanGovett has joined #openstack-nova | 11:49 | |
| openstackgerrit | jichenjc proposed openstack/nova: Cleanup HTTPRequest for security_groups test https://review.openstack.org/182630 | 11:52 |
| *** sdake has joined #openstack-nova | 11:54 | |
| *** dprince has joined #openstack-nova | 11:54 | |
| openstackgerrit | Hans Lindgren proposed openstack/nova: Remove unused refresh_security_group_members() call https://review.openstack.org/182632 | 11:55 |
| openstackgerrit | Hans Lindgren proposed openstack/nova: Make security group refresh with refresh_instance_security_groups() https://review.openstack.org/182633 | 11:55 |
| *** kiran-r has quit IRC | 11:55 | |
| *** tobe4333 has quit IRC | 11:57 | |
| Anticimex | johnthetubaguy: found my answer on http://docs.openstack.org/juno/config-reference/content/configuring-rpc.html i believe | 11:58 |
| Anticimex | johnthetubaguy: rabbit_hosts = $rabbit_host:$rabbit_port (ListOpt) RabbitMQ HA cluster host:port pairs. -- seems right? | 11:58 |
| *** IanGovett1 has joined #openstack-nova | 11:58 | |
| *** sdake has quit IRC | 11:59 | |
| *** igordcard_ has quit IRC | 11:59 | |
| *** igordcard_ has joined #openstack-nova | 11:59 | |
| *** sdake has joined #openstack-nova | 12:00 | |
| *** IanGovett has quit IRC | 12:01 | |
| bauzas | johnthetubaguy: I mean req-id | 12:01 |
| bauzas | johnthetubaguy: in the logs, you can see req-id but you have no way to see req-id in the instance object | 12:02 |
| bauzas | johnthetubaguy: just because it's the context request-id which is not persisted AFAICS | 12:02 |
| *** derekh_ has joined #openstack-nova | 12:08 | |
| *** derekh has quit IRC | 12:12 | |
| *** CongTo has quit IRC | 12:12 | |
| openstackgerrit | Lee Yarwood proposed openstack/nova: Multipath commands with error messages in stdout fail to parse https://review.openstack.org/172660 | 12:12 |
| *** aix has joined #openstack-nova | 12:13 | |
| *** signed8bit_ZZZzz has quit IRC | 12:14 | |
| johnthetubaguy | Anticimex: that sounds about right, we share the same oslo_messaging code as other projects, so hopefully its consistent between all the projects | 12:15 |
| johnthetubaguy | bauzas: not sure what you mean by "see req-id in the instance object"? | 12:15 |
| bauzas | johnthetubaguy: so, when looking at logs, we have the context provided for each line, right? | 12:17 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Ignore me. trying to log where urllib3 connectionpool is created https://review.openstack.org/182495 | 12:17 |
| bauzas | johnthetubaguy: something like req-edd46208-48f5-45cb-af7e-1f45b93b39b2 | 12:17 |
| johnthetubaguy | bauzas: right, for each API request you can track logs between all the different services, and between logs and notifications, and instance_actions in the DB also use request-id, for the ones that are tracked that way | 12:18 |
| bauzas | johnthetubaguy: right, my point is : request-id is not persisted when you nova show an instance | 12:18 |
| bauzas | johnthetubaguy: so, given you have an ERROR when booting an instance, you can see the last trace and the UUID | 12:19 |
| bauzas | johnthetubaguy: using nova show | 12:19 |
| johnthetubaguy | bauzas: its can't be, you can find some of them with nova list-instance-actions or something like that | 12:19 |
| bauzas | johnthetubaguy: IMHO having the req-id provided when doing a nova show would help operators to query the logs using this req-id | 12:20 |
| johnthetubaguy | bauzas: it can't. thats my point | 12:20 |
| bauzas | johnthetubaguy: it can't, but why ? | 12:20 |
| johnthetubaguy | bauzas: you could show the last request id, but there is a request id for every single API request you make | 12:20 |
| johnthetubaguy | bauzas: instance actions shows the related request ids for each instance | 12:21 |
| bauzas | johnthetubaguy: then I have to doublecheck | 12:21 |
| johnthetubaguy | bauzas: I am going to start with this stuff in the error doc, in the process of re-writing it all | 12:23 |
| bauzas | johnthetubaguy: oooooh I see now in the client, that's a contrib module | 12:24 |
| bauzas | johnthetubaguy: /os-instance-actions/ | 12:24 |
| * bauzas is just lost with all the extensions we have | 12:24 | |
| *** mwagner_lap has joined #openstack-nova | 12:25 | |
| *** Vek has joined #openstack-nova | 12:25 | |
| bauzas | johnthetubaguy: thanks for that, I see it's done http://developer.openstack.org/api-ref-compute-v2-ext.html | 12:25 |
| *** Jeffrey4l has quit IRC | 12:25 | |
| bauzas | johnthetubaguy: apologize for that, I never found it accessible | 12:26 |
| *** cbader has joined #openstack-nova | 12:26 | |
| *** vladikr has joined #openstack-nova | 12:26 | |
| *** otter768 has joined #openstack-nova | 12:26 | |
| johnthetubaguy | bauzas: its not really tested or complete, thats a different, but important issue! | 12:26 |
| bauzas | johnthetubaguy: yeah, since it was in the contib path for the client and only part of the extension for the API, I missed it | 12:27 |
| bauzas | johnthetubaguy: I just wonder that while it's not our duty to do that, we shouldn't provide a doc saying how to troubleshootr | 12:28 |
| Anticimex | johnthetubaguy: yeah, in http://www.severalnines.com/blog/high-availability-openstack-shoestring-budget-deploying-minimal-3-node-cluster they claim to configure it this way, so I assume it does work. will know in a bit | 12:28 |
| openstackgerrit | Tina Tang proposed openstack/nova: Fix in LibvirtISCSIVolumeDriver to avoid deleting incorrect device https://review.openstack.org/182565 | 12:28 |
| johnthetubaguy | Anticimex: there is a doc on this somewhere… http://docs.openstack.org/high-availability-guide/content/index.html | 12:29 |
| Anticimex | ah, crap. you are right | 12:30 |
| johnthetubaguy | bauzas: thats totally our duty | 12:30 |
| Anticimex | i've been in the HA doc several times | 12:30 |
| Anticimex | not sure how i have missed it | 12:30 |
| bauzas | johnthetubaguy: then, we can discuss that at the Ops-related session | 12:30 |
| johnthetubaguy | Anticimex: not sure if that has what you need, but we should patch that doc if its missing :) | 12:30 |
| bauzas | johnthetubaguy: because I'm sometimes lagging behind our API capabilities | 12:31 |
| *** dulek_ has quit IRC | 12:31 | |
| *** otter768 has quit IRC | 12:31 | |
| johnthetubaguy | bauzas: well, sadly is basically useless right now and needs fixing | 12:31 |
| Anticimex | johnthetubaguy: active/active, "Configure OpenStack services to use RabbitMQ" starts with "rabbit_hosts = ..." | 12:31 |
| bauzas | johnthetubaguy: and the API microversion should unleash the power | 12:31 |
| Anticimex | exactly what I was looking for | 12:31 |
| Anticimex | didn't turn up in google though :) | 12:31 |
| openstackgerrit | jichenjc proposed openstack/nova: Cleanup HTTPRequest for security_groups test https://review.openstack.org/182630 | 12:31 |
| johnthetubaguy | bauzas: we have tight tracking of the notifications which fils in the gaps quite well, thats the current way to deal with this | 12:31 |
| *** moshele has quit IRC | 12:31 | |
| johnthetubaguy | bauzas: yup | 12:32 |
| *** alexpilotti has joined #openstack-nova | 12:32 | |
| *** aysyd has joined #openstack-nova | 12:33 | |
| *** annegentle has joined #openstack-nova | 12:34 | |
| *** mpaolino has joined #openstack-nova | 12:34 | |
| *** Jeffrey4l has joined #openstack-nova | 12:39 | |
| *** salv-orlando has quit IRC | 12:39 | |
| *** matrohon has joined #openstack-nova | 12:41 | |
| *** dave-mccowan has joined #openstack-nova | 12:42 | |
| *** mestery has joined #openstack-nova | 12:44 | |
| *** READ10 has joined #openstack-nova | 12:44 | |
| *** moshele has joined #openstack-nova | 12:45 | |
| *** annegentle has quit IRC | 12:45 | |
| openstackgerrit | Noel Nelson Dsouza proposed openstack/nova: Nova can stop instance when Fake/Actual compute driver used https://review.openstack.org/182650 | 12:45 |
| *** annegentle has joined #openstack-nova | 12:46 | |
| *** gszasz has joined #openstack-nova | 12:47 | |
| *** cds has joined #openstack-nova | 12:48 | |
| cds | any nova devs free to look at https://bugs.launchpad.net/nova/+bug/1450291 ? blocking people from upgrading versions atm | 12:49 |
| openstack | Launchpad bug 1450291 in OpenStack Compute (nova) "nova-manage db sync not committing to database" [Undecided,New] | 12:49 |
| *** kiran-r has joined #openstack-nova | 12:51 | |
| johnthetubaguy | cds: I just marked it as critical, in the hope we can get some attention on that | 12:52 |
| cds | thank you, apprecite it :) | 12:52 |
| cds | appreciate* | 12:53 |
| *** tojuvone has quit IRC | 12:55 | |
| johnthetubaguy | mikal: dansmith: any ideas what could be happening here, given the test we do in turbo hipster? seems like nova-manage db sync fails: https://bugs.launchpad.net/nova/+bug/1450291 | 12:56 |
| openstack | Launchpad bug 1450291 in OpenStack Compute (nova) "nova-manage db sync not committing to database" [Critical,Confirmed] | 12:56 |
| johnthetubaguy | mriedem: this one ^ probably needs stable branch attention at somepoint, I tagged it for backport, I guess thats what you want to track these things? | 12:57 |
| openstackgerrit | Dmitry Guryanov proposed openstack/nova: libvirt: resize ploop disks with prl_disk_tool https://review.openstack.org/182390 | 13:00 |
| *** lucas-afk is now known as lucasagomes | 13:01 | |
| johnthetubaguy | cds: have you reproduced that bug at all, I wonder if its a config error thats hidden by the config logging looking a bit broken? | 13:01 |
| alex_xu | gilliard: I'm here | 13:03 |
| *** bkopilov_wfh has quit IRC | 13:04 | |
| *** baoli has joined #openstack-nova | 13:05 | |
| *** baoli has quit IRC | 13:05 | |
| *** nkrinner has quit IRC | 13:06 | |
| *** annegentle has quit IRC | 13:06 | |
| *** bkopilov_wfh has joined #openstack-nova | 13:07 | |
| alex_xu | johnthetubaguy: should I metion this https://review.openstack.org/127863 in api session etherpad? | 13:07 |
| *** annegentle has joined #openstack-nova | 13:07 | |
| openstackgerrit | Pranali Deore proposed openstack/nova: Unshelving volume backed instance fails https://review.openstack.org/170031 | 13:07 |
| *** baoli_ has joined #openstack-nova | 13:07 | |
| mriedem | johnthetubaguy: yeah, needs a fix on master first :) | 13:09 |
| johnthetubaguy | mriedem: totally | 13:09 |
| *** claudiub has joined #openstack-nova | 13:10 | |
| *** mriedem has quit IRC | 13:10 | |
| *** mriedem has joined #openstack-nova | 13:11 | |
| *** eglynn has quit IRC | 13:12 | |
| johnthetubaguy | alex_xu: this is the one for the API session, yes totally the sort of good stuff to add I feel: https://etherpad.openstack.org/p/YVR-nova-api-2.1-in-liberty | 13:12 |
| *** redbeard has joined #openstack-nova | 13:13 | |
| alex_xu | johnthetubaguy: ok, got it, thanks | 13:13 |
| *** baoli has joined #openstack-nova | 13:14 | |
| *** cds has quit IRC | 13:16 | |
| *** baoli_ has quit IRC | 13:16 | |
| claudiub | mriedem: hello. Can you take a look at a bugfix? You already did a review a couple of days ago, it already had a +2, it should be ready to go. :) https://review.openstack.org/#/c/181916/ | 13:21 |
| mriedem | claudiub: gate is a bit busted atm so i'm looking at that first | 13:22 |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: verify vCenter server certificate https://review.openstack.org/179728 | 13:22 |
| claudiub | mriedem: sure, take your time. :) | 13:22 |
| *** romainh has joined #openstack-nova | 13:23 | |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: Handle image size correctly for OVA and streamOptimized images https://review.openstack.org/168024 | 13:25 |
| johnthetubaguy | claudiub: I wonder if you could start a list of the top 3 to 5 most important patches that are ready for wider review, once you folks as a hyper-v subteam have done lots of reviews on the patches, in gerrit, and are happy with the code? | 13:25 |
| claudiub | johnthetubaguy: I'd be happy to. | 13:26 |
| *** sudipto has quit IRC | 13:27 | |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: Use virtual disk size instead of image size https://review.openstack.org/168013 | 13:27 |
| *** garyk has quit IRC | 13:29 | |
| claudiub | johnthetubaguy: IMO, we should put the commits in a line, so the conflicts are resolved and also it will show the preference and readyness of the commits. | 13:29 |
| johnthetubaguy | claudiub: I added a section at the end of the etherpad for you to start, I guess the above one can be the first patch: https://etherpad.openstack.org/p/liberty-nova-priorities-tracking | 13:30 |
| johnthetubaguy | claudiub: thats up to you, ideally we have the code refactored such that most patches never conflict | 13:30 |
| claudiub | johnthetubaguy: yeah, that's what I was saying, that the patch above should merge asap, since it fixes cold migration. | 13:30 |
| johnthetubaguy | claudiub: thats a great example | 13:30 |
| claudiub | johnthetubaguy: so, I'll do another review on our patches and if they look good to me, I'll add them to the etherpad list, in order of preference. | 13:32 |
| dansmith | johnthetubaguy: seems like maybe an ubuntu packaging thing, no? | 13:32 |
| johnthetubaguy | claudiub: just keep it to a top 3 or 5 patches though, if possible | 13:32 |
| johnthetubaguy | dansmith: oh really, I clearly didn't read that properly, sorry! | 13:32 |
| claudiub | johntheyubaguy: sure thing. | 13:32 |
| claudiub | tube* | 13:32 |
| claudiub | tuba* | 13:32 |
| johnthetubaguy | :) | 13:33 |
| johnthetubaguy | my nick was create such that IRC clients with autocomplete may one day rule the world | 13:33 |
| dansmith | johnthetubaguy: well, I dunno, just first glance seems like it might be | 13:33 |
| claudiub | i'm not good at musical instruments, not even at spelling them. :D | 13:33 |
| johnthetubaguy | dansmith: yeah, it feels like the nova config file might be in a non standard location, and not being passed into nova-manage, but I am not sure | 13:34 |
| dansmith | johnthetubaguy: same error with cinder db sync | 13:34 |
| *** exploreshaifali has joined #openstack-nova | 13:34 | |
| dansmith | johnthetubaguy: yeah, I dunno | 13:34 |
| johnthetubaguy | dansmith: although, someone said that did actually add the tables | 13:34 |
| dansmith | johnthetubaguy: kinda feel like it's not critical though, until someone on a different set of packages reports it | 13:34 |
| johnthetubaguy | dansmith: sorry, yeah, total over-reaction on my part | 13:34 |
| *** gszasz has quit IRC | 13:35 | |
| dansmith | johnthetubaguy: someone else said it doesn't happen on RDO | 13:35 |
| dansmith | but yeah, I dunno what emits that message, but it also doesn't fully seem to be a fatal error either | 13:36 |
| *** popw has quit IRC | 13:37 | |
| *** eglynn has joined #openstack-nova | 13:37 | |
| *** claudiub|2 has joined #openstack-nova | 13:38 | |
| johnthetubaguy | dansmith: yeah, I got the impression they didn't get any tables created, but maybe I miss read the other bits anyways | 13:38 |
| *** sdake_ has joined #openstack-nova | 13:38 | |
| dansmith | well, either way, if it's only happening on upgraded ubuntu packages, that surely seems like a deps problem or something | 13:39 |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Adds spec for instance live resize https://review.openstack.org/141219 | 13:39 |
| *** ankit_ag has quit IRC | 13:39 | |
| johnthetubaguy | dansmith: yeah, its just odd that the logs suggest the migrations all happened, thats why I was thinking it just getting written into the wrong DB, although that could be a dependency thing I guess (config not loading correctly, etc) | 13:40 |
| *** claudiub has quit IRC | 13:40 | |
| dansmith | yeah, that message seems like something not loading properly in our long chain of dependent things, I think | 13:40 |
| dansmith | but I'm not really sure who to ask to look at it | 13:40 |
| *** oro has quit IRC | 13:41 | |
| *** claudiub has joined #openstack-nova | 13:41 | |
| *** oro_ has quit IRC | 13:41 | |
| *** burt has joined #openstack-nova | 13:42 | |
| *** claudiub|3 has joined #openstack-nova | 13:42 | |
| *** sdake has quit IRC | 13:42 | |
| *** claudiub|2 has quit IRC | 13:42 | |
| johnthetubaguy | dansmith: me neither I am afraid, dims comes to mind as I see lots of oslo related messages | 13:43 |
| *** CongTo has joined #openstack-nova | 13:43 | |
| dansmith | yeah, that's probably a good call | 13:43 |
| *** claudiub|2 has joined #openstack-nova | 13:43 | |
| *** claudiub_ has joined #openstack-nova | 13:43 | |
| claudiub_ | johnthetubaguy: also, I've updated the instance-live-resize bp. There were plenty of comments, I hope I did not miss any. Also, I think there was a comment or 2 which was not clear enough to me. Whenever you have a little bit of time, please take a look. :) | 13:44 |
| *** claudiub|4 has joined #openstack-nova | 13:44 | |
| *** claudiu5 has joined #openstack-nova | 13:45 | |
| *** claudiu5 has quit IRC | 13:45 | |
| johnthetubaguy | claudiub: honestly a few days behind there right now, so working on the new stuff I have not yet touched first, but I will get back to that, but it could be post summit, although there is a summit session that includes online resize discussions BTW, should probably add a link to your spec in the etherpad | 13:45 |
| *** artom has quit IRC | 13:45 | |
| *** artom has joined #openstack-nova | 13:45 | |
| *** claudiub has quit IRC | 13:45 | |
| *** bjornar has quit IRC | 13:46 | |
| *** claudiub|3 has quit IRC | 13:46 | |
| *** BadCub-Hiding is now known as BadCub | 13:46 | |
| claudiub_ | johnthetubaguy: interesting. I'll have to make sure I get to that session | 13:47 |
| johnthetubaguy | claudiub_: https://bugs.launchpad.net/nova/+bug/1453835 it needs the priority setting on it, that would get it more reviews | 13:47 |
| openstack | Launchpad bug 1453835 in OpenStack Compute (nova) "Hyper-V: Nova cold resize / migration fails" [Medium,In progress] - Assigned to Claudiu Belu (cbelu) | 13:47 |
| johnthetubaguy | claudiub_: I just added the priorty | 13:47 |
| johnthetubaguy | claudiub_: can you describe on that bug why that regression was not picked up by the CI tests? | 13:48 |
| claudiub_ | johnthetubaguy: I do not have permission to set priority /.importance | 13:48 |
| *** claudiub|2 has quit IRC | 13:48 | |
| johnthetubaguy | claudiub_: you can give yourself permission, let me find the link | 13:48 |
| *** claudiub|4 has quit IRC | 13:48 | |
| *** links has quit IRC | 13:48 | |
| johnthetubaguy | claudiub_: https://wiki.openstack.org/wiki/Nova/BugTriage | 13:49 |
| johnthetubaguy | claudiub_: hmm, not sure that actually tell you... | 13:49 |
| johnthetubaguy | claudiub_: https://launchpad.net/~nova-bugs | 13:50 |
| claudiub_ | johnthetubaguy: I'm not sure why, it could be because the Hyper-V CI is setup with 2 compute nodes. I've hit the bug only having one compute node. We're currently testing Kilo on every system and this eventually came up. | 13:50 |
| johnthetubaguy | claudiub_: hmm, thats worth digging into, there maybe another bug there | 13:51 |
| *** moshele has quit IRC | 13:51 | |
| *** moshele has joined #openstack-nova | 13:52 | |
| *** dboik has joined #openstack-nova | 13:52 | |
| claudiub_ | johnthetubaguy: will do | 13:52 |
| claudiub_ | gtg! | 13:52 |
| johnthetubaguy | claudiub_: I just updated the wiki in the hope it helps others know about the bug team, does that look better? https://wiki.openstack.org/wiki/Nova/BugTriage | 13:53 |
| *** dboik_ has joined #openstack-nova | 13:53 | |
| *** claudiub_ has quit IRC | 13:57 | |
| *** dboik has quit IRC | 13:57 | |
| *** cfriesen has joined #openstack-nova | 13:57 | |
| *** markvoelker has quit IRC | 13:59 | |
| *** doron_afk has quit IRC | 14:00 | |
| *** sdake has joined #openstack-nova | 14:01 | |
| *** doron_afk has joined #openstack-nova | 14:01 | |
| *** mmedvede_ has quit IRC | 14:01 | |
| openstackgerrit | javeme proposed openstack/nova: VMware: Update to return the correct ESX iqn https://review.openstack.org/162700 | 14:01 |
| *** Kennan has quit IRC | 14:03 | |
| *** Kennan has joined #openstack-nova | 14:03 | |
| *** yingjun has joined #openstack-nova | 14:03 | |
| *** sdake_ has quit IRC | 14:04 | |
| *** amotoki_ has joined #openstack-nova | 14:06 | |
| *** mtanino has joined #openstack-nova | 14:07 | |
| dims | johnthetubaguy: dansmith: what seems to be the issue? | 14:08 |
| dansmith | dims: oslo. oslo is the issue. | 14:08 |
| * dansmith kids | 14:08 | |
| dansmith | dims: https://bugs.launchpad.net/nova/+bug/1450291 | 14:09 |
| openstack | Launchpad bug 1450291 in OpenStack Compute (nova) "nova-manage db sync not committing to database" [High,Incomplete] | 14:09 |
| dims | haha looking | 14:09 |
| dansmith | dims: thought you might be a good person to start with for figuring out what that error means | 14:09 |
| *** claudiub has joined #openstack-nova | 14:11 | |
| johnthetubaguy | dims: thanks for looking | 14:13 |
| *** eharney has quit IRC | 14:14 | |
| *** nickchase has joined #openstack-nova | 14:14 | |
| *** CongTo has quit IRC | 14:14 | |
| *** garyk has joined #openstack-nova | 14:14 | |
| *** nihilifer has quit IRC | 14:15 | |
| *** haomaiwang has joined #openstack-nova | 14:15 | |
| *** jaypipes has joined #openstack-nova | 14:15 | |
| mriedem | dansmith: log? | 14:16 |
| mriedem | dansmith: gate is f'ed btw | 14:16 |
| mriedem | grenade | 14:16 |
| dansmith | mriedem: log what? | 14:17 |
| mriedem | dansmith: you were saying something about oslo messages above | 14:17 |
| dansmith | mriedem: that bug is all I have | 14:18 |
| dansmith | [07:09:10] <dansmith> dims: https://bugs.launchpad.net/nova/+bug/1450291 | 14:18 |
| openstack | Launchpad bug 1450291 in OpenStack Compute (nova) "nova-manage db sync not committing to database" [High,Incomplete] | 14:18 |
| mriedem | oh | 14:18 |
| mriedem | maybe ubuntu threw in some nova db migrations of their own :{P | 14:18 |
| dims | johnthetubaguy: dansmith: "No handlers could be found for logger" just means the code path that ended with a LOG.xyz did not end up configuring logging, so there's something screwy in their environment or how they installed stuff. have any of us been able to replicate the problem? | 14:18 |
| dansmith | mriedem: someone said they didn't have the problem with RDO, and all the others are ubuntu 14.04 or whatever, so kinda seems like ubuntu deps | 14:18 |
| dansmith | mriedem: heh, I don't think so :) | 14:19 |
| dansmith | dims: that explains why the migrations still get applied | 14:19 |
| dansmith | dims: like I said above, seemed more like a warning than anything | 14:19 |
| dims | right | 14:19 |
| *** claudiub|2 has joined #openstack-nova | 14:19 | |
| dansmith | dims: so that's cool, probably just a config issue then? | 14:19 |
| dims | yep | 14:19 |
| mriedem | need to know the version of oslo.db, sqlalchemy-migrate and sqlalchemy when they hit this | 14:20 |
| mriedem | added that to the bug | 14:20 |
| dims | thanks mriedem | 14:20 |
| dansmith | is this in any way our deal though? | 14:20 |
| *** claudiub|3 has joined #openstack-nova | 14:20 | |
| mriedem | me? | 14:20 |
| dansmith | meaning, | 14:20 |
| mriedem | hells naw | 14:20 |
| dansmith | is this purely ubuntu packaging? | 14:20 |
| mriedem | idk | 14:20 |
| mriedem | my packages are just fine :) | 14:20 |
| dansmith | I just think we should probably drop the severity based on the impact, especially since I don't think we have anything to do here | 14:21 |
| mriedem | would be good to get the versions from their site-packages when it's running and failing though | 14:21 |
| dansmith | I'm not even sure who still works for canonical to ask | 14:21 |
| *** claudiub|4 has joined #openstack-nova | 14:21 | |
| mriedem | i know some distros like to munge the requirements and specify their own | 14:21 |
| dansmith | I thought all distros did that? :) | 14:21 |
| *** tonytan4ever has joined #openstack-nova | 14:21 | |
| mriedem | like rdo zero's out requirements.txt in the rpmbuild and assumes it's rpm requirements are golden | 14:22 |
| *** mpaolino has quit IRC | 14:22 | |
| mriedem | we dont | 14:22 |
| *** claudiub has quit IRC | 14:22 | |
| dansmith | well, that's kindof to be expected I think :) | 14:22 |
| mriedem | we're masochists, we keep our deps up to the same version as what's in requirements.txt | 14:22 |
| *** claudiub has joined #openstack-nova | 14:22 | |
| mriedem | which means like daily package updates during development | 14:22 |
| dansmith | right, but we support things shipped via rpm, | 14:22 |
| dansmith | so fi we don't do that, things can shift because pip decides to "help" | 14:23 |
| mriedem | we're using rpms too | 14:23 |
| mriedem | not pip | 14:23 |
| mriedem | but we base the requirements in our rpm spec on what's in requirements.txt | 14:23 |
| *** claudiu7 has joined #openstack-nova | 14:23 | |
| *** claudiub|2 has quit IRC | 14:23 | |
| dansmith | right, but we have other things that openstack in our distro, | 14:23 |
| mriedem | so when g-r updates that, we update our packages if needed | 14:23 |
| mriedem | it's painful | 14:23 |
| dansmith | so we have to make sure we have a common package that works for everything | 14:24 |
| *** jecarey has joined #openstack-nova | 14:24 | |
| *** haomaiwang has quit IRC | 14:25 | |
| dansmith | mriedem: anyway, you said grenade is effing up the gate this morning? | 14:25 |
| *** claudiub_ has joined #openstack-nova | 14:25 | |
| *** claudiub|3 has quit IRC | 14:25 | |
| mriedem | well, https://bugs.launchpad.net/python-openstackclient/+bug/1454467 | 14:25 |
| openstack | Launchpad bug 1454467 in python-openstackclient "Latest version 1.0.4 doesn't work with master" [Undecided,New] | 14:25 |
| mriedem | looks like jogo saved all of our asses https://review.openstack.org/#/c/182524/ | 14:26 |
| *** bkopilov has joined #openstack-nova | 14:26 | |
| *** claudiub|4 has quit IRC | 14:26 | |
| *** claudiu7 has quit IRC | 14:26 | |
| *** dave-mccowan has quit IRC | 14:26 | |
| *** dulek has joined #openstack-nova | 14:26 | |
| *** claudiub has quit IRC | 14:27 | |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: use get_datastore_by_ref from oslo.vmware https://review.openstack.org/163043 | 14:27 |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: use get_object_properties_dict from oslo.vmware https://review.openstack.org/163831 | 14:27 |
| *** otter768 has joined #openstack-nova | 14:27 | |
| *** bkopilov_wfh has quit IRC | 14:27 | |
| dansmith | fun | 14:28 |
| *** afazekas has joined #openstack-nova | 14:29 | |
| *** atuvenie has quit IRC | 14:31 | |
| *** otter768 has quit IRC | 14:32 | |
| *** pixelbeat has joined #openstack-nova | 14:32 | |
| *** pixelbeat_ has quit IRC | 14:34 | |
| mikal | . | 14:37 |
| *** cds has joined #openstack-nova | 14:37 | |
| mikal | johnthetubaguy: you still want to talk TH? Or is that done now? | 14:37 |
| *** lennyb has joined #openstack-nova | 14:37 | |
| johnthetubaguy | mikal: TH? | 14:37 |
| johnthetubaguy | mikal: oh turbo hipster | 14:38 |
| mriedem | cds: what does nova-manage db version show? | 14:38 |
| mriedem | 'nova-manage db version' | 14:38 |
| *** dave-mccowan has joined #openstack-nova | 14:39 | |
| cds | No handlers could be found for logger "oslo_config.cfg" 280 | 14:39 |
| mriedem | that's it? | 14:39 |
| cds | yes, 280 is on a newline though | 14:39 |
| *** tsekiyama has joined #openstack-nova | 14:39 | |
| dansmith | mikal: I think we've got a handle on it | 14:40 |
| mriedem | cds: then you're upgraded | 14:40 |
| dansmith | mikal: johnthetubaguy was concerned db_sync was completely broken | 14:40 |
| mriedem | https://github.com/openstack/nova/tree/stable/kilo/nova/db/sqlalchemy/migrate_repo/versions 280 is the latest kilo migration | 14:40 |
| johnthetubaguy | yeah, I over reacted | 14:40 |
| mikal | dansmith: oh, ok | 14:41 |
| mriedem | cds: what version of oslo.config is installed? | 14:41 |
| mikal | So, just noting that TH doesn't test postgres at all because we've never found a user with a production database who is willing to share | 14:41 |
| johnthetubaguy | cds: do you have nova.conf correct on node you are running nova-manage? | 14:41 |
| cds | i dumped my config at https://bugs.launchpad.net/nova/+bug/1450291 | 14:41 |
| openstack | Launchpad bug 1450291 in OpenStack Compute (nova) "nova-manage db sync not committing to database" [High,Incomplete] | 14:41 |
| mriedem | johnthetubaguy: he's at migration 280 | 14:41 |
| mriedem | so the migrations ran | 14:41 |
| dansmith | yeah, this is just a logging-unconfigured warning I think | 14:42 |
| dansmith | which is what I first thought, | 14:42 |
| dansmith | but dims confirmed | 14:42 |
| dansmith | and we know the migrations are happening, so.. | 14:42 |
| johnthetubaguy | ah, OK | 14:42 |
| johnthetubaguy | cds: does the DB have the correct information in it? | 14:42 |
| *** yingjun has quit IRC | 14:42 | |
| *** oro has joined #openstack-nova | 14:42 | |
| cds | python-oslo.config 1:1.9.3-0ubuntu1~cloud0 | 14:42 |
| cds | yes, but nova doesn't seem to find it | 14:42 |
| *** oro has quit IRC | 14:42 | |
| cds | my instances table remains | 14:42 |
| johnthetubaguy | cds: what is the nova error you are seeing? | 14:43 |
| cds | but nova lost comes up blank | 14:43 |
| cds | list* | 14:43 |
| mriedem | but there are instances in the instances table? | 14:43 |
| dansmith | sure sounds like some config mismatch | 14:43 |
| johnthetubaguy | dansmith: +1 | 14:44 |
| *** mestery has quit IRC | 14:44 | |
| *** claudiub_ has quit IRC | 14:45 | |
| openstackgerrit | Zhengguang Ou proposed openstack/nova: Fix the race condition between detaching volume and creating image https://review.openstack.org/166152 | 14:47 |
| *** zsais has joined #openstack-nova | 14:48 | |
| *** Murali has quit IRC | 14:48 | |
| *** mpaolino has joined #openstack-nova | 14:49 | |
| *** cds has quit IRC | 14:50 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: improve instance names on VC https://review.openstack.org/166608 | 14:50 |
| *** patrickeast has joined #openstack-nova | 14:52 | |
| *** hemna_ is now known as hemna | 14:53 | |
| *** xgsa has quit IRC | 14:54 | |
| *** garyk has quit IRC | 14:56 | |
| *** garyk has joined #openstack-nova | 14:56 | |
| *** josecastroleon has quit IRC | 14:57 | |
| *** bkopilov has quit IRC | 14:57 | |
| *** kmartin has joined #openstack-nova | 14:59 | |
| cfriesen | discovered what I think is a day-one race in cpu pinning. https://bugs.launchpad.net/nova/+bug/1454451 Who are the right people to contact. ndipanov I assume, anyone else? | 14:59 |
| openstack | Launchpad bug 1454451 in OpenStack Compute (nova) "simultaneous boot of multiple instances leads to cpu pinning overlap" [Undecided,New] - Assigned to Chris Friesen (cbf123) | 14:59 |
| *** Marga_ has joined #openstack-nova | 14:59 | |
| dansmith | ndipanov is out for a while | 15:01 |
| dansmith | cfriesen: danpb or sahid probably | 15:02 |
| *** marcusvrn_ has joined #openstack-nova | 15:02 | |
| *** Sukhdev has joined #openstack-nova | 15:02 | |
| *** Marga_ has quit IRC | 15:03 | |
| *** Marga_ has joined #openstack-nova | 15:03 | |
| danpb | i'm not familiar enough with the schedling code really - sahid or nikola when he returns in a week & 1/2 | 15:04 |
| cfriesen | not a race in the scheduler...resource tracking in nova-compute | 15:05 |
| *** b3nt_pin is now known as beagles | 15:05 | |
| danpb | sure, i consider that to be part of scheduling :-) | 15:05 |
| openstackgerrit | Victor Stinner proposed openstack/nova: Replace MySQL-python with PyMySQL in tests https://review.openstack.org/182709 | 15:07 |
| *** jichen has quit IRC | 15:07 | |
| haypo | dansmith, hi. i rewrote my patch for VersionedObjectDictCompat: https://review.openstack.org/#/c/182699/ | 15:09 |
| dansmith | haypo: yeah, why did you do that? | 15:09 |
| dansmith | haypo: in general, it's much preferred to re-use your existing review instead of abandoning and moving to another one | 15:09 |
| kashyap | cfriesen: I can give it a shot to reproduce sometime this week, since I recall testing similar stuff from git some months ago. | 15:09 |
| dansmith | haypo: I really really hate the if statement in the middle of the class, FYI | 15:10 |
| haypo | dansmith, ah. i splitted my change into two changes, i removed notes in a first change | 15:10 |
| dansmith | haypo: still, you shouldn't have abandoned the original | 15:10 |
| cfriesen | kashyap: sweet, thanks. | 15:10 |
| dansmith | haypo: now the history is confused | 15:10 |
| *** unicell1 has joined #openstack-nova | 15:10 | |
| kashyap | cfriesen: Tomm is a holiday here, so I may try it Friday. | 15:10 |
| haypo | dansmith, ok sorry, i will reuse the same change id next time | 15:11 |
| *** bkopilov has joined #openstack-nova | 15:11 | |
| dansmith | thanks | 15:11 |
| *** n0ano has quit IRC | 15:11 | |
| haypo | dansmith, do you prefer testing the python major version inside the function? i chose to test in the class body is to not add performance overhead at each code, the class body is only executed once | 15:11 |
| *** unicell has quit IRC | 15:12 | |
| *** apuimedo has quit IRC | 15:12 | |
| dansmith | haypo: I'll reply on the review in a bit | 15:12 |
| *** nelsnelson has joined #openstack-nova | 15:14 | |
| *** garyk has quit IRC | 15:15 | |
| *** garyk has joined #openstack-nova | 15:15 | |
| *** dims has quit IRC | 15:15 | |
| *** dims has joined #openstack-nova | 15:16 | |
| *** moshele has quit IRC | 15:16 | |
| *** matrohon has quit IRC | 15:16 | |
| *** garyk has quit IRC | 15:19 | |
| *** tojuvone has joined #openstack-nova | 15:19 | |
| *** bfic has quit IRC | 15:19 | |
| *** jistr is now known as jistr|mtg | 15:19 | |
| *** salv-orlando has joined #openstack-nova | 15:22 | |
| openstackgerrit | Andrew Laski proposed openstack/nova-specs: Cells: Add host mapping table and object https://review.openstack.org/182715 | 15:24 |
| markus_z | If anyone is able to establish a secure websocket connection (wss) with nova-serialproxy to the guests, please contact me. I can't figure it out. | 15:24 |
| *** afazekas_ has quit IRC | 15:24 | |
| *** afazekas has quit IRC | 15:24 | |
| *** Nikolay_St has quit IRC | 15:26 | |
| *** Nikolay_St has joined #openstack-nova | 15:27 | |
| mriedem | directxman12: ^? | 15:27 |
| gilliard | sahid wrote a howto I think? | 15:28 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova: WIP switch to using os-brick https://review.openstack.org/175569 | 15:28 |
| *** jistr|mtg is now known as jistr | 15:29 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Eventlet supports python logging directly https://review.openstack.org/179801 | 15:29 |
| *** sdake has quit IRC | 15:30 | |
| *** sdake has joined #openstack-nova | 15:30 | |
| *** moshele has joined #openstack-nova | 15:32 | |
| *** yamahata has joined #openstack-nova | 15:33 | |
| *** ildikov has quit IRC | 15:33 | |
| *** browne has joined #openstack-nova | 15:34 | |
| *** thangp has joined #openstack-nova | 15:35 | |
| openstackgerrit | dave-mcnally proposed openstack/nova: Add policy to cover snapshotting of volume backed instances https://review.openstack.org/176805 | 15:36 |
| *** yamahata has quit IRC | 15:36 | |
| *** Nikolay_St has quit IRC | 15:36 | |
| *** yamahata has joined #openstack-nova | 15:36 | |
| *** mriedem has quit IRC | 15:37 | |
| *** mriedem has joined #openstack-nova | 15:37 | |
| *** exploreshaifali has quit IRC | 15:37 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Add rootwrap daemon mode support https://review.openstack.org/180695 | 15:39 |
| *** salv-orlando has quit IRC | 15:41 | |
| *** lennyb has quit IRC | 15:43 | |
| *** bkopilov has quit IRC | 15:44 | |
| *** eharney has joined #openstack-nova | 15:45 | |
| *** bkopilov has joined #openstack-nova | 15:45 | |
| *** obedmr has joined #openstack-nova | 15:46 | |
| sahid | gilliard: yes, markus_z you can find some informationon devref manual | 15:46 |
| markus_z | directxman12: sahid: I've found how to start the nova-serialproxy. But this didn't cover how the setup should be when "wss" instead of "ws" should be used. | 15:48 |
| dansmith | wss is like https | 15:48 |
| dansmith | ws = cleartext, wss = ssl'd | 15:48 |
| openstackgerrit | Joel Coffman proposed openstack/nova: libvirt: Clean up unit tests for _hard_reboot https://review.openstack.org/177398 | 15:49 |
| markus_z | dansmith: right, and "websockify" is not able to create a ssl connection to my compute node | 15:49 |
| tojuvone | Was there a discussion reserved in summit for service groups and tooz? Referring the dev mailinglist discussion on "Service group foundations and features". | 15:49 |
| dansmith | markus_z: ah, I see | 15:50 |
| markus_z | I assume that the nova-serialproxy needs a key-file or a cert-file. This is not described in the documentation. | 15:51 |
| *** shakamunyi has quit IRC | 15:51 | |
| *** barra204 has quit IRC | 15:51 | |
| markus_z | Devstack uses HTTP with WS and there it works fine. But now I want to create a "real-live" environment with HTTPS and WSS. | 15:52 |
| lucasagomes | hey quick q. When we issue a "nova delete" in one instance that still "spawning" it doesn't invoke the destroy() method in the driver right? | 15:53 |
| tojuvone | or should I reserve a slot for having API changes for external monitoring system to update Nova states like in https://review.openstack.org/169836 ? | 15:54 |
| *** kiran-r has quit IRC | 15:54 | |
| lucasagomes | cause in Ironic we are able to stop provisioning at some points mid-operation, and it seems that "nova delete" for not invoking destroy() is cause the instance to get stuck | 15:54 |
| mriedem | lucasagomes: it's a crap shoot | 15:55 |
| dansmith | lucasagomes: looks like it should still call it to me | 15:56 |
| lucasagomes | mriedem, right, I mean you can still look at the ironic node and issue a delete in ironic and then delete in nova and the instance goes away | 15:57 |
| dansmith | but I can see something else being half-setup causing it to bail early | 15:57 |
| lucasagomes | dansmith, yeah... it's because we need to boot a machine so sometimes it the deployment can fail due some network problems | 15:58 |
| *** dprince has quit IRC | 15:58 | |
| lucasagomes | and the ramdisk can't connect to the machine etc | 15:58 |
| lucasagomes | it will timeout eventually | 15:58 |
| lucasagomes | but sometimes you just want to get rid of that instance immediately and free the ne | 15:58 |
| lucasagomes | node* | 15:58 |
| dansmith | lucasagomes: I meant something in nova preventing the delete from ever actually calling driver.delete | 15:58 |
| lucasagomes | dansmith, yeah, it seems so. I added some debug in the destroy() method and it's not being invoked | 15:59 |
| markus_z | sahid: If you could give me any pointer you have about websockets and nova-serialproxy in this mail http://lists.openstack.org/pipermail/openstack/2015-May/012661.html that would be really great! I have to leave for today. | 15:59 |
| lucasagomes | dansmith, I will investigate more. Just asked to see if someone had some idea off the top of their heads | 15:59 |
| *** markus_z has quit IRC | 16:00 | |
| *** vilobhmm1 has joined #openstack-nova | 16:00 | |
| *** redbeard has quit IRC | 16:01 | |
| obedmr | hi there, does this message sound familiar to someone "RTNETLINK answers: File exists" ? It's happening when starting 5 VMs in a row, 4 succeed and 1 failed | 16:03 |
| *** unicell1 has quit IRC | 16:06 | |
| *** obedmr has quit IRC | 16:06 | |
| *** jlanoux has quit IRC | 16:07 | |
| *** obedmr has joined #openstack-nova | 16:07 | |
| *** dulek has quit IRC | 16:08 | |
| *** garyk has joined #openstack-nova | 16:08 | |
| *** jcoufal has quit IRC | 16:08 | |
| mriedem | obedmr: goolge | 16:08 |
| mriedem | 16:09 | |
| mriedem | https://bugs.launchpad.net/nova/+bug/1000580 | 16:09 |
| openstack | Launchpad bug 1000580 in neutron "Quantum Linux Bridge interface driver plug operation fails with RTNETLINK error" [High,Fix released] - Assigned to Sumit Naiksatam (snaiksat) | 16:09 |
| mriedem | however, that's super old | 16:09 |
| mriedem | obedmr: are you on diablo? | 16:09 |
| mriedem | or essex | 16:09 |
| obedmr | mriedem: actually I saw those links, but the thing is that I'm on Kilo | 16:11 |
| *** dsanders has joined #openstack-nova | 16:11 | |
| obedmr | using nova-network legacy | 16:11 |
| *** erlon has joined #openstack-nova | 16:12 | |
| mriedem | obedmr: report a new bug with your configuration details, version you're running, and a paste of the trace | 16:12 |
| *** rwsu has joined #openstack-nova | 16:13 | |
| obedmr | mriedem: sure, doing it, thanks | 16:13 |
| openstackgerrit | Qin Zhao proposed openstack/nova: Make resize api of compute manager to send object https://review.openstack.org/161860 | 16:15 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Add rootwrap daemon mode support https://review.openstack.org/180695 | 16:19 |
| openstackgerrit | Victor Stinner proposed openstack/nova: Run tests with PyMySQL on Python 3 https://review.openstack.org/182709 | 16:20 |
| *** xyhuang has joined #openstack-nova | 16:21 | |
| *** lpetrut has quit IRC | 16:21 | |
| *** Nic has joined #openstack-nova | 16:21 | |
| *** alanf-mc has joined #openstack-nova | 16:24 | |
| *** vilobhmm1 has quit IRC | 16:25 | |
| *** nihilifer has joined #openstack-nova | 16:25 | |
| *** jistr has quit IRC | 16:25 | |
| *** otter768 has joined #openstack-nova | 16:28 | |
| *** sahid has quit IRC | 16:28 | |
| *** nihilifer has quit IRC | 16:28 | |
| *** alanf-mc has quit IRC | 16:28 | |
| *** dprince has joined #openstack-nova | 16:29 | |
| *** haypo has quit IRC | 16:32 | |
| *** otter768 has quit IRC | 16:32 | |
| *** alanf-mc has joined #openstack-nova | 16:33 | |
| _gryf | dansmith, hi | 16:34 |
| dansmith | hello | 16:34 |
| _gryf | dansmith, I saw your comments | 16:34 |
| _gryf | dansmith, thanks | 16:34 |
| *** tonytan4ever has quit IRC | 16:35 | |
| _gryf | dansmith, there is a little difference between how it is implemented right now from what you are proposing | 16:35 |
| _gryf | dansmith, https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/plugins/v3/services.py#L107 | 16:35 |
| _gryf | dansmith, this line show exactly how id field is used in case of services api | 16:36 |
| _gryf | its exactly the same way as it was before in 2.0 version of the API | 16:36 |
| _gryf | dansmith, so the req lacks of {service: .. } part | 16:37 |
| *** derekh_ has quit IRC | 16:37 | |
| _gryf | dansmith, and the id is used differently one can expect :) | 16:37 |
| *** alanf-mc_ has joined #openstack-nova | 16:38 | |
| dansmith | _gryf: are you talking about the mark-service-down spec? | 16:38 |
| _gryf | dansmith, yes | 16:38 |
| *** nihilifer has joined #openstack-nova | 16:38 | |
| dansmith | _gryf: I'm proposing that the spec behave like the current api does, instead of something different | 16:39 |
| *** unicell has joined #openstack-nova | 16:39 | |
| _gryf | dansmith, let me repeat myself: currently, for v2.1 api, os-services are behave like I've described | 16:40 |
| *** alanf-mc has quit IRC | 16:41 | |
| *** Jeffrey4l has quit IRC | 16:41 | |
| _gryf | dansmith, so IMHO there are 3 opts we have: either we make all the calls for os-services coherent in correct way, make it look the same as the current way, ot make this particular call like you proposed | 16:41 |
| *** mpaolino has quit IRC | 16:42 | |
| dansmith | _gryf: I see now, based on the code how incredibly broken that api is (if id == "enable" <-- insane) | 16:42 |
| _gryf | dansmith, YES! | 16:42 |
| _gryf | :) | 16:42 |
| *** jlanoux has joined #openstack-nova | 16:42 | |
| *** arnaud____ has joined #openstack-nova | 16:43 | |
| _gryf | dansmith, i feel that right way to do is report a bug for broken api and make the new call in correct way | 16:44 |
| dansmith | _gryf: there's no way to make this really correct without adding a new thing, so we might as well just continue to mirror what is there and hopefully implement a new services endpoint that is less stupid later | 16:45 |
| *** marun has joined #openstack-nova | 16:46 | |
| _gryf | dansmith, hm. why cannot we just report a bug for it? 2.1 is still current, isn't it? | 16:46 |
| *** claudiub has joined #openstack-nova | 16:46 | |
| dansmith | we can't change it, we can only replace it | 16:46 |
| *** amotoki_ has quit IRC | 16:46 | |
| _gryf | oh | 16:47 |
| _gryf | that suck | 16:47 |
| dansmith | well, you know, stable APIs and all.. | 16:47 |
| _gryf | right... | 16:47 |
| cfriesen | dansmith: change it with a microversion? or is that too much of a change? | 16:48 |
| dansmith | cfriesen: we still have to honor the old mechanism until we drop v2.0 support (which may never happen) | 16:49 |
| cfriesen | dansmith: I guess we'd have to add a new endpoint with a microversion | 16:49 |
| dansmith | cfriesen: the microversion would be "add a new /services-that-does-not-suck-as-much" endpoint | 16:49 |
| dansmith | right | 16:49 |
| _gryf | oh crap | 16:49 |
| dansmith | I would strongly recommend you not place that road block in front of landing this spec :) | 16:49 |
| *** arnaud____ has quit IRC | 16:49 | |
| cfriesen | dansmith: but it's so simple. :) | 16:50 |
| _gryf | dansmith, but i like your endpoint name ;) | 16:50 |
| *** ppal has quit IRC | 16:50 | |
| *** browne has quit IRC | 16:51 | |
| *** romainh has left #openstack-nova | 16:51 | |
| *** patrickeast_ has joined #openstack-nova | 16:51 | |
| *** lpetrut has joined #openstack-nova | 16:52 | |
| openstackgerrit | javeme proposed openstack/nova: VMware: Prefer ShutdownGuest to PowerOffVM_Task https://review.openstack.org/181634 | 16:52 |
| *** patrickeast has quit IRC | 16:52 | |
| *** patrickeast_ is now known as patrickeast | 16:52 | |
| *** signed8bit has joined #openstack-nova | 16:55 | |
| *** Marga_ has quit IRC | 16:57 | |
| openstackgerrit | Jean-Daniel Bonnetot proposed openstack/nova-specs: Sort instances if possible inside an host aggregate https://review.openstack.org/182758 | 16:58 |
| *** vilobhmm1 has joined #openstack-nova | 16:59 | |
| *** sdake has quit IRC | 17:00 | |
| *** jcoufal has joined #openstack-nova | 17:01 | |
| *** tjones1 has joined #openstack-nova | 17:02 | |
| *** Sukhdev has quit IRC | 17:02 | |
| *** jcoufal has quit IRC | 17:02 | |
| *** Sukhdev has joined #openstack-nova | 17:03 | |
| *** patrickeast has quit IRC | 17:03 | |
| *** ijw has joined #openstack-nova | 17:04 | |
| *** davideag_ has joined #openstack-nova | 17:04 | |
| *** winston-d is now known as winston-d_zzZ | 17:04 | |
| jogo | sdague: when is the last time you ran abandon_old_reviews? | 17:05 |
| *** zzzeek has joined #openstack-nova | 17:06 | |
| *** davideagnello has quit IRC | 17:07 | |
| lucasagomes | dansmith, hi there, the delete is getting stuck in this lock here https://github.com/openstack/nova/blob/3841b8254e0dd820d26377312b6c004f4bd9b58e/nova/compute/manager.py#L2431 | 17:08 |
| lucasagomes | just a FYI, I don't know yet how I'm going to resolve it | 17:08 |
| dansmith | ah | 17:08 |
| openstackgerrit | Eric Harney proposed openstack/nova: libvirt: Proxy RBD library call into a separate thread https://review.openstack.org/180212 | 17:09 |
| dansmith | is the build process waiting on ironic? | 17:09 |
| dansmith | lucasagomes: ^ | 17:09 |
| lucasagomes | dansmith, probably. Lemme investigate it | 17:09 |
| dansmith | seems likely | 17:09 |
| lucasagomes | dansmith, thanks for the tip! | 17:10 |
| *** Nic1 has joined #openstack-nova | 17:10 | |
| *** Nic has quit IRC | 17:10 | |
| dansmith | np | 17:10 |
| jogo | looks like its been a while | 17:11 |
| * jogo runs abandon old reivews | 17:11 | |
| *** alanf-mc_ has quit IRC | 17:12 | |
| *** davideagnello has joined #openstack-nova | 17:12 | |
| lucasagomes | dansmith, yeah it looks like spawn() is waiting in a loopingcall waiting the node to get ACTIVE (in ironic) | 17:13 |
| *** moshele has quit IRC | 17:13 | |
| *** alanf-mc has joined #openstack-nova | 17:13 | |
| lucasagomes | and it doesn't return until it's ACTIVE or ERROR (in case of a timeout) | 17:13 |
| *** claudiub|2 has joined #openstack-nova | 17:13 | |
| lucasagomes | perhaps somehow we need to cleanup the lock when it's waiting somehow | 17:13 |
| directxman12 | markus_z: /win 13 | 17:14 |
| *** achanda has joined #openstack-nova | 17:14 | |
| openstackgerrit | javeme proposed openstack/nova: VMware: Prefer ShutdownGuest to PowerOffVM_Task https://review.openstack.org/181634 | 17:14 |
| *** claudiub has quit IRC | 17:14 | |
| dansmith | lucasagomes: well, it's doing what it's supposed to.. does ironic not give us any indication that there has been a failure so that it can break out? | 17:15 |
| *** davideag_ has quit IRC | 17:15 | |
| dansmith | jaypipes: when you're free, vilobhmm1 and I had trouble verifying something about that computefilter call discussion we had the other day | 17:16 |
| jaypipes | dansmith: yes? | 17:16 |
| dansmith | this: https://github.com/openstack/nova/blob/master/nova/scheduler/driver.py#L51 | 17:16 |
| dansmith | jaypipes: seems to only ever be called by the chance scheduler | 17:16 |
| lucasagomes | dansmith, yeah it does. If something fails in ironic the node goes to DEPLOYFAIL provison_state and then the loop is nova is interrupted | 17:16 |
| dansmith | jaypipes: the filter scheduler gets its list of hosts from ComputeNodeList | 17:16 |
| lucasagomes | and instance gets marked as ERROR | 17:16 |
| lucasagomes | it's just that inconvinience that we can't delete an instance with Ironic mid-deployment | 17:17 |
| lucasagomes | via nova | 17:17 |
| lucasagomes | it's supported via ironic api directly | 17:17 |
| lucasagomes | anyway... I will see if there's anything we can do in the nova side tomorrow. If not we better document it | 17:17 |
| lucasagomes | dansmith, cheers for the help! | 17:17 |
| dansmith | lucasagomes: probably one of those situations where having a driver that calls another distributed service is a sticky place to be | 17:18 |
| dansmith | lucasagomes: but yep, cool | 17:18 |
| lucasagomes | dansmith, yeah. Also deployment in Ironic happens in two phases, before the ramdisk gets into the disk | 17:18 |
| lucasagomes | and after it gets in the disk | 17:18 |
| dansmith | jaypipes: filter scheduler gets it from here I think: https://github.com/openstack/nova/blob/master/nova/scheduler/host_manager.py#L535 | 17:18 |
| *** nihilifer has quit IRC | 17:18 | |
| lucasagomes | so there's synchronization from nova -> ironic -> ramdisk (and back) | 17:18 |
| *** vilobhmm1 has quit IRC | 17:18 | |
| lucasagomes | it's not easy :-) | 17:18 |
| dansmith | lucasagomes: yeah | 17:18 |
| lucasagomes | ramdisk get into the node* | 17:19 |
| *** vilobhmm1 has joined #openstack-nova | 17:19 | |
| jaypipes | dansmith: well, no, it gets it from here: https://github.com/openstack/nova/blob/master/nova/scheduler/host_manager.py#L532 | 17:19 |
| *** mpavone has quit IRC | 17:19 | |
| dansmith | lucasagomes: we could have some state I guess to allow it to be interrupted, but right now it's stateless, so.. :/ | 17:19 |
| jaypipes | dansmith: and that has the disabled=False hard-coded into it in the DB API layer. | 17:20 |
| dansmith | jaypipes: well, yeah | 17:20 |
| *** harlowja has quit IRC | 17:20 | |
| dansmith | jaypipes: it's the hosts_up() thing we were wondering about | 17:20 |
| jaypipes | dansmith: with this: https://github.com/openstack/nova/blob/master/nova/scheduler/host_manager.py#L185-L187 | 17:20 |
| *** harlowja has joined #openstack-nova | 17:20 | |
| dansmith | jaypipes: I think because he was expecting that to be called because you said it was | 17:20 |
| jaypipes | dansmith: which is the equivalent :) | 17:20 |
| jaypipes | dansmith: it's a mess for sure. | 17:20 |
| lucasagomes | yeah, well, let's see I don't know the right fix yet but at least I know where/why it's failing | 17:21 |
| dansmith | jogo: you're blowing up my review box | 17:21 |
| dansmith | jaypipes: I'm not saying it's different, I just want to make sure I'm not missing how hosts_up() actually gets called from filter scheduler | 17:21 |
| jogo | dansmith: running sdague's abandoner | 17:22 |
| jogo | it should abandon 185 or so | 17:22 |
| jogo | we have a lot of dead reviews | 17:23 |
| jogo | had | 17:23 |
| dansmith | jogo: yeah, it's just annoying me with a mail ding about every second | 17:23 |
| jogo | hehe | 17:23 |
| openstackgerrit | Chris Friesen proposed openstack/nova: Fix race between resource audit and cpu pinning https://review.openstack.org/182766 | 17:23 |
| jogo | I hope you don't get an audible ping or something | 17:23 |
| cfriesen | would love to see comments on the direction taken in this fix ^ | 17:23 |
| cfriesen | sort of nasty to be adding items to the conductor instance_update | 17:24 |
| *** claudiub has joined #openstack-nova | 17:24 | |
| dansmith | cfriesen: yeah, can't do that, IMHO | 17:25 |
| *** lucasagomes is now known as lucas-dinner | 17:25 | |
| *** sudipto has joined #openstack-nova | 17:25 | |
| *** claudiub|2 has quit IRC | 17:26 | |
| cfriesen | dansmith: the thing is I'm not convinced we can call instance.save() since that would persist all the changes since the last save, whereas the old code was only updating the node/host at that point | 17:27 |
| cfriesen | dansmith: but we need to save the numa_topology claim while under the semaphore | 17:28 |
| bauzas | does anyone know how to access a client manager which is in the contrib path of the novaclient ? | 17:28 |
| bauzas | ie. https://github.com/openstack/python-novaclient/blob/master/novaclient/v2/contrib/migrations.py ? | 17:28 |
| cfriesen | bauzas: would love to see your thoughts on the fix above and the bug related to it | 17:28 |
| bauzas | cfriesen: unfortunately, I'm quite off upstream since I have to close some internal things before I go to Vancouver | 17:29 |
| cfriesen | bauzas: no worries | 17:30 |
| cfriesen | bauzas: me too...this is one of my internal things. :) | 17:30 |
| *** salv-orlando has joined #openstack-nova | 17:30 | |
| bauzas | and since tomorrow is (again, ergh) a French holiday, tonight is my last day | 17:30 |
| dansmith | cfriesen: passing that value to instance_update() doesn't even make sense right? | 17:30 |
| dansmith | cfriesen: that goes straight to the db update call, which knows nothing about instance['numa_topology'] AFAICT | 17:30 |
| *** fawadkhaliq has joined #openstack-nova | 17:31 | |
| cfriesen | dansmith: well, it's not blowing up my lab. :) checking | 17:32 |
| dansmith | cfriesen: should be ignored, I think | 17:32 |
| mriedem | jogo: this seems to be the kind of thing you're usually interested in https://review.openstack.org/#/c/181944/ | 17:33 |
| *** browne has joined #openstack-nova | 17:33 | |
| openstackgerrit | Mikhail Feoktistov proposed openstack/nova: Virtuozzo instance resize support https://review.openstack.org/182257 | 17:33 |
| *** 17SACKCIT has joined #openstack-nova | 17:34 | |
| *** 7YUAAALBR has joined #openstack-nova | 17:34 | |
| jogo | mriedem: yup! | 17:34 |
| *** thangp has quit IRC | 17:34 | |
| jogo | very slick | 17:34 |
| *** 7YUAAALBR has quit IRC | 17:34 | |
| *** 17SACKCIT has quit IRC | 17:34 | |
| bauzas | melwitt: around for a question ? | 17:35 |
| bauzas | melwitt: I need your novaclient expertise :) | 17:35 |
| *** asselin has joined #openstack-nova | 17:35 | |
| cfriesen | dansmith: I added 'numa_topology' to "allowed_updates" in the conductor. or are you saying that the DB API won't do anything with it? | 17:35 |
| openstackgerrit | Mikhail Feoktistov proposed openstack/nova: Virtuozzo instance resize support https://review.openstack.org/182257 | 17:35 |
| dansmith | cfriesen: the db | 17:35 |
| melwitt | bauzas: checking | 17:36 |
| dansmith | cfriesen: when you call instance_update, you're just flat updating the db model, which doesn't have a numa_topology on the instance | 17:36 |
| bauzas | melwitt: what I basically wonder is how to call a nova.migrations.list() method when just instanciating the client object directly | 17:36 |
| *** thangp has joined #openstack-nova | 17:37 | |
| dansmith | cfriesen: and since it's stored in the DB as an object, if you send a non-object you'll break the copy in the db, and if you send an object, you'll be breaking rpc | 17:37 |
| bauzas | https://github.com/openstack/python-novaclient/blob/master/novaclient/v2/contrib/migrations.py | 17:37 |
| dansmith | cfriesen: it's actually stored in instance_extra.numa_topology | 17:37 |
| vilobhmm1 | dansmith, jaypipes : sorry got caught up in some meeting | 17:37 |
| *** nihilifer has joined #openstack-nova | 17:37 | |
| vilobhmm1 | but dansmith yes jaypipes is right | 17:37 |
| bauzas | melwitt: for some reason, the CLI has some way to call cs.migrations.list() but my local object doesn't know it | 17:37 |
| dansmith | vilobhmm1: about the service query filtering for disabled=True, right? | 17:38 |
| bauzas | melwitt: so I guess I have to extend the client | 17:38 |
| vilobhmm1 | yes | 17:38 |
| dansmith | vilobhmm1: right, I'm not arguing that point :) | 17:38 |
| bauzas | melwitt: but since I'm not a client expert, I'm lost in translation | 17:38 |
| *** xyhuang has quit IRC | 17:38 | |
| vilobhmm1 | for the host_up thing | 17:38 |
| *** doude has quit IRC | 17:38 | |
| *** nihilifer has quit IRC | 17:38 | |
| melwitt | bauzas: I wrote a similar example in here https://bugs.launchpad.net/python-novaclient/+bug/1313266 you have to import the extension, put it in a list, pass the list to the client object, then you can access it | 17:39 |
| openstack | Launchpad bug 1313266 in python-keystoneclient "List extensions not exposed in the user-facing library API" [Wishlist,Fix released] - Assigned to Matt Fischer (mfisch) | 17:39 |
| *** xyhuang has joined #openstack-nova | 17:39 | |
| bauzas | melwitt: lemme look | 17:39 |
| bauzas | melwitt: I see, introspecting the object and extending it live ? | 17:39 |
| *** yamahata has quit IRC | 17:40 | |
| bauzas | melwitt: probably that https://bugs.launchpad.net/python-novaclient/+bug/1313266/comments/9 ? | 17:40 |
| openstack | Launchpad bug 1313266 in python-keystoneclient "List extensions not exposed in the user-facing library API" [Wishlist,Fix released] - Assigned to Matt Fischer (mfisch) | 17:40 |
| melwitt | bauzas: yes | 17:41 |
| dansmith | cfriesen: I think you might want/need to wait until the RT is fixed for instance objects | 17:41 |
| *** marcusvrn_ is now known as marcusvrn | 17:41 | |
| bauzas | melwitt: ok, will see how to do this properly | 17:41 |
| *** belmoreira has joined #openstack-nova | 17:41 | |
| bauzas | melwitt: I owe you a beer/soft/soda/whatever in Vancouver | 17:41 |
| *** alexpilotti has quit IRC | 17:41 | |
| melwitt | heh :) | 17:42 |
| dansmith | ah, beers.. the oft-promised, never-collected currency of opensource | 17:42 |
| bauzas | dansmith: :) | 17:42 |
| bauzas | dansmith: that's why we have evening events right ? | 17:42 |
| bauzas | beers are fre | 17:42 |
| bauzas | free | 17:43 |
| dansmith | heh, resets the counts, right? | 17:43 |
| bauzas | :) | 17:43 |
| melwitt | I can't even remember how I figured out the client thing. it must be in the code somewhere. I should add it to the novaclient doc | 17:43 |
| bauzas | melwitt: the worst thing is that the extensions parameter is not provided in the main client | 17:43 |
| *** belmoreira has quit IRC | 17:43 | |
| *** salv-orlando has quit IRC | 17:44 | |
| cfriesen | dansmith: just confirmed that it's not working, DB not being updated. I'll have to figure out something though.\ | 17:44 |
| bauzas | ie. when you correctly do nova.client(version, ...) | 17:44 |
| dansmith | cfriesen: good, because if it was working, I was going to be very concerned :) | 17:44 |
| bauzas | melwitt: so I will probably monkeypatch the attribute | 17:44 |
| bauzas | melwitt: oh sec, there is a kwargs | 17:44 |
| bauzas | marvelous | 17:45 |
| cfriesen | dansmith: how about something truly disgusting...read the instance object from the DB, update the affected fields, call instance.save()? | 17:45 |
| cfriesen | not as a final fix, but a workaround | 17:45 |
| bauzas | melwitt: yeah, love that https://github.com/openstack/python-novaclient/blob/master/novaclient/v2/contrib/migrations.py | 17:46 |
| dansmith | cfriesen: ask some RT people. Personally, I think that's likely to generate too much load, especially given that the audience for this feature is small | 17:46 |
| dansmith | cfriesen: I'd much prefer to live with the possibility until we can fix it properly | 17:46 |
| cfriesen | dansmith: I'm not even thinking of submitting that, just as an internal workaround till we get a proper upstream fix. | 17:46 |
| dansmith | cfriesen: oh sure | 17:46 |
| dansmith | cfriesen: damn, was hoping to get another -2 out of that :P | 17:46 |
| dansmith | cfriesen: does the scheduler pick which cpus you get pinned to? If so, I'm not sure I understand how this fixes any race | 17:47 |
| dansmith | er, if not I meant | 17:48 |
| openstackgerrit | Matt Riedemann proposed openstack/nova: neutron: log hypervisor_macs before raising PortNotUsable https://review.openstack.org/175458 | 17:48 |
| *** danpb has quit IRC | 17:49 | |
| *** patrickeast has joined #openstack-nova | 17:50 | |
| cfriesen | dansmith: the instance claim picks the cpus that we get pinned to, and stores them in the numa topology. currently the claimed numa topology gets updated in the instance shortly after releasing the semaphore | 17:51 |
| cfriesen | we need to update it in the instance in the DB before releasing the semaphore | 17:51 |
| cfriesen | dansmith: because the resource audit takes the semaphore and then queries the DB for all the instances on this host | 17:52 |
| dansmith | cfriesen: so the claim process grabs the semaphore and reads the current pinnings from the db before assigning a pinning ? | 17:52 |
| bauzas | melwitt: thanks a lot, it works | 17:52 |
| cfriesen | dansmith: ResourceTracker.instance_claim() takes the semaphore and calls down into claim, which looks at the current pinnings stored locally in nova-compute | 17:53 |
| dansmith | cfriesen: okay, that's what I thought, that's why I don't understand why we have to write to the DB with the semaphore held | 17:54 |
| cfriesen | dansmith: but the resource audit runs every minute and rebuilds the local picture of in-use resources based on what's in the DB | 17:54 |
| melwitt | bauzas: whew, I'm glad. also found how I realized it, this is what the shell does to find all the extensions and pass them in the kwarg so everything works https://github.com/openstack/python-novaclient/blob/master/novaclient/shell.py#L525-L527 | 17:54 |
| dansmith | cfriesen: ah, okay | 17:54 |
| dansmith | cfriesen: so you're racing with *that* | 17:54 |
| bauzas | melwitt: yeah | 17:54 |
| cfriesen | dansmith: yes | 17:54 |
| bauzas | melwitt: I also have to beat with the host-evacuate logic, but I can just mimic it since nova.evacuate() is there | 17:55 |
| cfriesen | dansmith: so currently we can claim a cpu for an instance, run the resource audit, then claim the same cpu for another instance | 17:55 |
| bauzas | melwitt: (all of that is not upstream, don't be worried about that :p ) | 17:55 |
| dansmith | cfriesen: yeah, I get it | 17:55 |
| *** nihilifer has joined #openstack-nova | 17:56 | |
| openstackgerrit | John Garbutt proposed openstack/nova: devref: discuss current state of error handling https://review.openstack.org/178635 | 17:57 |
| *** haomaiwa_ has joined #openstack-nova | 17:57 | |
| openstackgerrit | Andrew Laski proposed openstack/nova: WIP Retry a cell delete if host constraint fails https://review.openstack.org/182772 | 17:58 |
| *** fawadk has joined #openstack-nova | 17:58 | |
| *** marcusvrn has quit IRC | 17:59 | |
| *** doron_afk has quit IRC | 17:59 | |
| openstackgerrit | Erickson Filipe Guedes dos Santos proposed openstack/nova: Create column allocated in Quota table https://review.openstack.org/151327 | 18:00 |
| *** jcook- has joined #openstack-nova | 18:01 | |
| *** fawadkhaliq has quit IRC | 18:01 | |
| *** Sukhdev_ has joined #openstack-nova | 18:03 | |
| *** jcook has quit IRC | 18:04 | |
| *** jcook- is now known as jcook | 18:04 | |
| *** annegentle has quit IRC | 18:04 | |
| *** fawadkhaliq has joined #openstack-nova | 18:04 | |
| *** Sukhdev has quit IRC | 18:05 | |
| *** yamahata has joined #openstack-nova | 18:05 | |
| *** fawadk has quit IRC | 18:06 | |
| *** ijw has quit IRC | 18:07 | |
| *** ijw has joined #openstack-nova | 18:07 | |
| *** cds has joined #openstack-nova | 18:09 | |
| cds | anyone have time to look at https://bugs.launchpad.net/nova/+bug/1446082 ? | 18:10 |
| openstack | Launchpad bug 1446082 in OpenStack Compute (nova) "Instance without "extra" data crashes nova-compute" [Undecided,In progress] - Assigned to Christoph Dwertmann (cdwertmann) | 18:10 |
| *** ijw_ has joined #openstack-nova | 18:10 | |
| *** ajo has quit IRC | 18:10 | |
| *** fawadk has joined #openstack-nova | 18:11 | |
| *** ijw has quit IRC | 18:12 | |
| *** luv_ has joined #openstack-nova | 18:12 | |
| *** alanf-mc has quit IRC | 18:12 | |
| mriedem | dansmith: hrm ^ | 18:12 |
| dansmith | will look ina sec | 18:13 |
| *** luv has quit IRC | 18:13 | |
| *** luv_ is now known as luv | 18:13 | |
| *** romainh has joined #openstack-nova | 18:13 | |
| mriedem | cds: so you ran migrate_flavor_data ? | 18:13 |
| *** nihilifer has quit IRC | 18:13 | |
| *** fawadkhaliq has quit IRC | 18:13 | |
| cds | i have not, did not know that existed | 18:14 |
| cds | what is the command? | 18:14 |
| *** fawadk has quit IRC | 18:14 | |
| mriedem | sec | 18:14 |
| mriedem | https://wiki.openstack.org/wiki/ReleaseNotes/Kilo#Upgrade_Support | 18:14 |
| mriedem | https://wiki.openstack.org/wiki/ReleaseNotes/Kilo#Upgrade_Notes_2 | 18:14 |
| mriedem | After fully upgrading to kilo (i.e. all nodes are running kilo code), you should start a background migration of flavor information from its old home to its new home. Kilo conductor nodes will do this on the fly when necessary, but the rest of the idle data needs to be migrated in the the background. This is critical to complete before the Liberty release, where support for the old location will be dropped. Use "nova-manage migrate-flavo | 18:15 |
| *** gokrokve has joined #openstack-nova | 18:16 | |
| *** alanf-mc has joined #openstack-nova | 18:17 | |
| mriedem | but according to https://github.com/openstack/nova/commit/bf8b3946e6964059dce581f54ed4f928e2cdf5ab that's just something that has to be run before upgrading to liberty | 18:18 |
| cds | hmm | 18:19 |
| cds | let me try this nova-manage db migrate_flavor_data | 18:19 |
| cds | %(total)i instances matched query, %(done)i completed {'total': 0, 'done': 0} | 18:20 |
| cds | compute node still dies | 18:20 |
| cds | err %(total)i instances matched query, %(done)i completed {'total': 1, 'done': 1} | 18:20 |
| dansmith | cds: you know you can't upgrade from icehouse to kilo directly right? | 18:20 |
| *** pixelbeat has quit IRC | 18:20 | |
| cds | correct | 18:20 |
| cds | i am on juno | 18:20 |
| mriedem | dansmith: the original bug reporter was on icehouse | 18:21 |
| dansmith | oh, I see | 18:21 |
| mriedem | https://github.com/openstack/nova/commit/6d3c920a569ff843dde3463eb07ab04bd1847e7f | 18:21 |
| mriedem | so the 252 migration creates the instance_extra table but doesn't move any data into it, i'm assuming that is supposed to be a live schema upgrade / data migration in the objects on first access? | 18:23 |
| dansmith | the trace in the original bug doesn't make much sense, | 18:23 |
| dansmith | because it implies that the instance itself is None | 18:23 |
| dansmith | mriedem: yeah | 18:24 |
| *** Nikolay_St has joined #openstack-nova | 18:24 | |
| dansmith | that was in juno though right? | 18:24 |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova-specs: Allow admins to query user keypairs and create new on their behalf https://review.openstack.org/175579 | 18:24 |
| mriedem | dansmith: yeah https://review.openstack.org/#/c/108097/ | 18:24 |
| mriedem | so...upgrade from icehouse directly to kilo w/o migrating any data, run kilo code which assumes the data migration happened in juno? | 18:25 |
| cds | i guess it's possible i upgraded from icehouse->juno and never ran the db migration | 18:26 |
| dansmith | no, because we didn't have anything from i->j to fill in extra, AFAIK | 18:26 |
| *** tojuvone has quit IRC | 18:26 | |
| cds | but i was on juno before installing kilo | 18:26 |
| dansmith | and we couldn't have required them to touch all their instances | 18:26 |
| dansmith | cds: you abandoned your patch because why again? | 18:26 |
| cds | sorry, that wasn't my patch | 18:27 |
| mriedem | dansmith: the guy screwed up | 18:27 |
| mriedem | dansmith: he pushed https://review.openstack.org/#/c/175288/ and then abandoned that for change 2, which Vek said do it in the first | 18:28 |
| mriedem | so now both are abandoned :) | 18:28 |
| dansmith | right, then the other was abandoned | 18:28 |
| *** Marga_ has joined #openstack-nova | 18:28 | |
| dansmith | cds: that's not you? | 18:28 |
| mriedem | derp | 18:28 |
| mriedem | no | 18:28 |
| mriedem | chris != craig | 18:28 |
| dansmith | okay | 18:28 |
| cds | :) | 18:28 |
| * mriedem acts like he knows all of these guys | 18:28 | |
| dansmith | ISTR a patch to create instance_extra in certain cases | 18:28 |
| dansmith | lemme look for a sec | 18:29 |
| *** otter768 has joined #openstack-nova | 18:29 | |
| *** tonytan4ever has joined #openstack-nova | 18:29 | |
| dansmith | I'm just not sure how this ever passed grenade if it was broken | 18:31 |
| openstackgerrit | Qin Zhao proposed openstack/nova: Make resize api of compute manager to send flavor object https://review.openstack.org/161860 | 18:32 |
| mriedem | dansmith: do we actually put numa stuff in extra? | 18:32 |
| mriedem | in grnade | 18:32 |
| dansmith | mriedem: doesn't matter I don't think | 18:32 |
| dansmith | unless we don't query with that in expected_attrs if that filter isn't enabled, but I doubt it | 18:32 |
| *** otter768 has quit IRC | 18:33 | |
| cds | any suggestions on what i could do in the meantime? is it possible to roll back nova to juno, rerun db sync, and the move back to kilo? | 18:33 |
| dansmith | oh, you know why | 18:33 |
| dansmith | when we added this, we were doing independent queries | 18:34 |
| dansmith | and then we probably optimized after we moved past the point where new instances get extra records by default | 18:34 |
| *** claudiub|2 has joined #openstack-nova | 18:34 | |
| *** claudiub has quit IRC | 18:34 | |
| *** ijw_ has quit IRC | 18:34 | |
| *** Marga_ has quit IRC | 18:35 | |
| *** ijw has joined #openstack-nova | 18:35 | |
| *** Marga_ has joined #openstack-nova | 18:35 | |
| *** ijw has quit IRC | 18:35 | |
| dansmith | so since those patches are abandonded, I think I'll just write a new one | 18:35 |
| dansmith | wow, too many D's | 18:35 |
| dansmith | cds: can you test a patch for me? | 18:35 |
| cds | of course | 18:36 |
| *** ijw has joined #openstack-nova | 18:36 | |
| dansmith | okay, give me a bit | 18:36 |
| *** jlanoux_ has joined #openstack-nova | 18:36 | |
| cds | let me rejoin on something other than webchat | 18:37 |
| *** jlanoux has quit IRC | 18:37 | |
| *** cds has quit IRC | 18:37 | |
| *** cds has joined #openstack-nova | 18:37 | |
| cds | alright | 18:37 |
| dansmith | zzzeek: I think I've asked this a hundred times, but what's the newer way to tell if a thing is loaded in a sqla object? | 18:37 |
| dansmith | without triggering a load | 18:38 |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova-specs: Allow admins to query user keypairs and create new on their behalf https://review.openstack.org/175579 | 18:38 |
| zzzeek | dansmith: do you mean an attribute, like myobject.some_attribute | 18:38 |
| dansmith | I thought it changed between 0.8 and 0.9 | 18:38 |
| dansmith | zzzeek: yes | 18:38 |
| zzzeek | dansmith: the always-worked way is just to look in __dict__, myobject.__dict__[‘some_attribute’] | 18:38 |
| dansmith | zzzeek: so "if 'some_attribute' in obj.__dict__" to detect presence? | 18:39 |
| zzzeek | dansmith: the newish way is probably inspect(myobject).attrs.some_attribute .loaded_value | 18:39 |
| zzzeek | dansmith: if the value is there, then that’s the value. the more modern non-hacky way is the inspect() form | 18:39 |
| dansmith | okay, probably should do the old way since this is for a backport | 18:39 |
| dansmith | zzzeek: thanks | 18:39 |
| zzzeek | dansmith: good plan | 18:39 |
| *** Nikolay_St has quit IRC | 18:43 | |
| *** pixelbeat has joined #openstack-nova | 18:45 | |
| *** sudipto has quit IRC | 18:45 | |
| *** claudiub has joined #openstack-nova | 18:46 | |
| *** aix has quit IRC | 18:47 | |
| openstackgerrit | sajeesh cs proposed openstack/nova: Nested Quota Driver https://review.openstack.org/149828 | 18:48 |
| *** claudiub|2 has quit IRC | 18:50 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Fix loading things in instance_extra for old instances https://review.openstack.org/182787 | 18:50 |
| dansmith | cds: ^ | 18:51 |
| *** Sukhdev_ has quit IRC | 18:51 | |
| cds | thanks | 18:51 |
| dansmith | cds: looks like it should still apply to kilo | 18:51 |
| cds | ok | 18:51 |
| dansmith | now, the migrate_flavor_data thing really should be creating extra records | 18:51 |
| dansmith | you said you tried that and didn't get any, is that right? | 18:52 |
| cds | i got one | 18:52 |
| dansmith | one? | 18:52 |
| cds | nova-manage db migrate_flavor_data 50 | 18:52 |
| cds | No handlers could be found for logger "oslo_config.cfg" | 18:52 |
| cds | %(total)i instances matched query, %(done)i completed {'total': 1, 'done': 1} | 18:52 |
| dansmith | oh I see, missed that second one | 18:52 |
| dansmith | well, that's concerning | 18:53 |
| dansmith | cds: presumably you have lots of undeleted instances in your instances table, right? | 18:53 |
| dansmith | cds: and do you have some extra rows, or just the one it created? | 18:53 |
| *** yapeng has joined #openstack-nova | 18:54 | |
| *** xyhuang has quit IRC | 18:57 | |
| *** baoli has quit IRC | 18:57 | |
| cds | sec | 19:00 |
| cds | applying patch | 19:00 |
| *** annegentle has joined #openstack-nova | 19:00 | |
| *** Longgeek has quit IRC | 19:01 | |
| *** Longgeek has joined #openstack-nova | 19:01 | |
| *** Longgeek has quit IRC | 19:02 | |
| *** Longgeek has joined #openstack-nova | 19:02 | |
| dansmith | writing tests now, | 19:02 |
| dansmith | and replicated the behavior, which is fixed with that patch | 19:02 |
| *** yapeng has quit IRC | 19:02 | |
| mriedem | dansmith: ping me when the tests are up, looked good otherwise | 19:02 |
| dansmith | k | 19:03 |
| *** claudiub|2 has joined #openstack-nova | 19:03 | |
| *** Longgeek has quit IRC | 19:04 | |
| *** Longgeek has joined #openstack-nova | 19:04 | |
| *** claudiub has quit IRC | 19:05 | |
| *** exploreshaifali has joined #openstack-nova | 19:05 | |
| *** sdake has joined #openstack-nova | 19:06 | |
| cds | dansmith: besides applying the patch and removing the old .pyc file, why is error pointing back to stale code? | 19:06 |
| mriedem | dansmith: melwitt was seeing the same issues in the change that i was at the same time :) | 19:07 |
| mriedem | i had to look closer | 19:07 |
| dansmith | cds: meaning you still see the error but it's pointing to a line in the code that is no longer right? | 19:08 |
| cds | correct | 19:08 |
| dansmith | mriedem: which change? | 19:08 |
| mriedem | dansmith: ps1 | 19:08 |
| dansmith | cds: sure seems like a pyc thing | 19:08 |
| mriedem | https://review.openstack.org/#/c/182787/ comments inline | 19:08 |
| cds | same error, same line, but that line no longer exists in the source file | 19:08 |
| melwitt | cds: did you restart the service? | 19:09 |
| melwitt | or is it a nova-manage thing | 19:09 |
| mriedem | nova-manage already ran | 19:09 |
| openstackgerrit | Joe Gordon proposed openstack/nova-specs: Make backlog usable https://review.openstack.org/182793 | 19:09 |
| mriedem | has to patch, remove pyc, and restart nova-compute | 19:09 |
| jogo | turns out nova-specs backlog wasn't actually usable until now ^ | 19:10 |
| *** pixelbeat has quit IRC | 19:12 | |
| dansmith | mriedem: melwitt yeah, did that flavor change prospectively, didn't actually test anything yet, good catch | 19:12 |
| *** alanf-mc has quit IRC | 19:13 | |
| *** alanf-mc has joined #openstack-nova | 19:14 | |
| cds | dansmith: got it running with the updated patch, same thing unfortunately | 19:15 |
| cds | line 515 in your patch | 19:15 |
| dansmith | cds: can you paste the trace? | 19:15 |
| dansmith | pastebin * | 19:15 |
| cds | sure | 19:15 |
| cds | http://pastebin.com/fip6YBTG | 19:15 |
| *** alanf-mc has quit IRC | 19:15 | |
| *** exploreshaifali has quit IRC | 19:16 | |
| *** alanf-mc has joined #openstack-nova | 19:16 | |
| *** Sukhdev has joined #openstack-nova | 19:16 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: detach cinder volume when instance destroyed https://review.openstack.org/169732 | 19:16 |
| openstackgerrit | garyk proposed openstack/nova: VMware: remove redundant check for block devices https://review.openstack.org/170641 | 19:16 |
| mriedem | hmm, db_inst.get('extra') returning None? | 19:18 |
| dansmith | I'm wondering if it's actually in the object, but None | 19:18 |
| dansmith | either way, foo.get().get() makes no sense | 19:19 |
| cds | i can get my db for certain fields if it would help | 19:19 |
| cds | check my db* | 19:19 |
| dansmith | this is code semantics | 19:19 |
| mriedem | cds: well, you're getting into a block where the extra column/field is on the instance record | 19:19 |
| dansmith | sdague: | 19:22 |
| dansmith | er | 19:22 |
| dansmith | cds: | 19:22 |
| openstackgerrit | Dan Smith proposed openstack/nova: Fix loading things in instance_extra for old instances https://review.openstack.org/182787 | 19:22 |
| dansmith | don't need the test parts of course | 19:22 |
| mriedem | looks better | 19:23 |
| dansmith | adding test for the flavor part | 19:24 |
| openstackgerrit | Marian Horban proposed openstack/nova: Disassociating of floating IPs corrected https://review.openstack.org/149358 | 19:24 |
| *** claudiub has joined #openstack-nova | 19:25 | |
| cds | woahhhh | 19:25 |
| cds | it worked :) | 19:25 |
| cds | afaik at least | 19:25 |
| *** alanf-mc has quit IRC | 19:25 | |
| dansmith | cds: did we warn you about the hourly rate for this kind of support? :) | 19:25 |
| cds | lol | 19:26 |
| *** Nikolay_St has joined #openstack-nova | 19:26 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Fix loading things in instance_extra for old instances https://review.openstack.org/182787 | 19:26 |
| dansmith | updated to include coverage for flavor in the test: | 19:26 |
| cds | so tell me this.. am i going to run into more problems? | 19:26 |
| dansmith | ^ | 19:26 |
| *** alanf-mc has joined #openstack-nova | 19:27 | |
| cds | should i snapshot these VMs and recreate new instances? | 19:27 |
| dansmith | cds: well, that question is very, uh, loaded :) | 19:27 |
| dansmith | cds: but no, you shouldn't need to do that | 19:27 |
| *** claudiub|2 has quit IRC | 19:27 | |
| dansmith | cds: we just goofed a little on two-cycle-old instances here I think | 19:28 |
| cds | hmm | 19:28 |
| dansmith | I haven't run git-blame yet to see if it was my fault or not, but I assume so.. | 19:28 |
| cds | so liberty is released, i should be able to db sync and be alright? | 19:29 |
| dansmith | you mean *when* it's released? | 19:29 |
| cds | yes | 19:29 |
| dansmith | yes, and I'll backport this fix so that when your packages are next updated, it should have this in there | 19:30 |
| cds | awesome | 19:30 |
| dansmith | cds: the only thing that has me still concerned is why you were only seeing one instance get hit with the migrate_flavor_data thing | 19:30 |
| dansmith | because that will be a requirement before liberty will let you do db_sync | 19:30 |
| *** belmoreira has joined #openstack-nova | 19:32 | |
| cds | hmm | 19:32 |
| cds | could i recreate all the flavors now that i am on kilo and redeploy the vms? | 19:33 |
| cds | this is a dev cluster with <50 vms | 19:33 |
| *** belmoreira has quit IRC | 19:33 | |
| dansmith | this is actually not about flavors, but about instances and how they reference the flavors | 19:33 |
| cds | ah | 19:33 |
| dansmith | so anything you do to touch the instance will do the migration for you, | 19:33 |
| *** dave-mccowan has quit IRC | 19:33 | |
| dansmith | the migrate_flavor_data thing is just to migrate any instances that aren't touched to trigger that migration | 19:33 |
| dansmith | could you run it again with this patch applied? maybe this was preventing it from hitting everything | 19:34 |
| cds | sure | 19:34 |
| dansmith | I kinda expect not given that we don't catch and ignore this in the migration thing... | 19:35 |
| cds | yeah | 19:35 |
| cds | still 0 | 19:35 |
| *** frickler has quit IRC | 19:36 | |
| dansmith | it's super odd that you didn't see this error when trying to do that in the first place actually | 19:36 |
| *** nkrinner has joined #openstack-nova | 19:36 | |
| cds | any suggestions? or just hope it gets caught and fixed upstream | 19:38 |
| dansmith | I'm thinking and looking at the code... | 19:38 |
| *** eharney has quit IRC | 19:38 | |
| *** pece has quit IRC | 19:39 | |
| dansmith | could you maybe do some queries in your database to count things? | 19:39 |
| openstackgerrit | Joe Gordon proposed openstack/nova-specs: Make backlog usable https://review.openstack.org/182793 | 19:39 |
| dansmith | basically, it would be helpful to know how many non-deleted instances you have that don't have a corresponding record in instance_extra | 19:39 |
| cds | of course | 19:40 |
| *** jlanoux_ has quit IRC | 19:40 | |
| *** packet has joined #openstack-nova | 19:40 | |
| dansmith | zzzeek: https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L6011-6013 | 19:40 |
| dansmith | zzzeek: if we do that and InstanceExtra doesn't exist for that instance, will we get a record? | 19:41 |
| zzzeek | dansmith: yes | 19:41 |
| cfriesen | dansmith: back to that race in nova-compute...looks like I can call objects.InstanceNUMATopology.get_by_instance_uuid() and update just that object rather than the whole instance. | 19:41 |
| dansmith | hmm, okay | 19:41 |
| zzzeek | dansmith: 3050 points, can’t go wrong: http://stackoverflow.com/a/38578/34549 | 19:42 |
| dansmith | zzzeek: yeah, I need to understand that better, sorry | 19:43 |
| zzzeek | dansmith: not a problem | 19:45 |
| * zzzeek did SQL for like ~3 years not knowing JOIN | 19:45 | |
| dims | lol :) | 19:45 |
| dansmith | even still, the ORM between me and SQL often makes me second guess things | 19:47 |
| dansmith | because I just assume it's pure magic | 19:47 |
| *** Sukhdev_ has joined #openstack-nova | 19:47 | |
| *** jwcroppe has joined #openstack-nova | 19:48 | |
| melwitt | it's alchemy :) | 19:48 |
| *** IanGovett has joined #openstack-nova | 19:48 | |
| dansmith | right | 19:48 |
| dansmith | if it was named SQLnotscaryatallcomeonin then it'd be different :) | 19:48 |
| *** ijw_ has joined #openstack-nova | 19:49 | |
| melwitt | that's for sure! | 19:49 |
| dansmith | that's about the only thing harder to type quickly than sqlalchemy too | 19:49 |
| *** IanGovett1 has quit IRC | 19:50 | |
| *** Sukhdev has quit IRC | 19:50 | |
| *** jwcroppe_ has quit IRC | 19:50 | |
| *** arnaud____ has joined #openstack-nova | 19:50 | |
| cds | dansmith: 33 instance uuids in instance_extras | 19:50 |
| cds | 37 non-deleted VMs | 19:50 |
| cds | in instances | 19:50 |
| *** ijw has quit IRC | 19:51 | |
| *** erlon has quit IRC | 19:51 | |
| dansmith | hmm | 19:52 |
| bauzas | mikal: as per https://bugs.launchpad.net/nova/+bug/1450798 it seems that's an Invalid one | 19:52 |
| openstack | Launchpad bug 1450798 in OpenStack Compute (nova) "Multiple command injection vulns in schema_diff tool" [Undecided,New] | 19:52 |
| bauzas | mikal: you agree ? | 19:53 |
| dansmith | cds: you're checking non-deletedness how? | 19:54 |
| cds | select deleted from instances where deleted = 0; | 19:54 |
| dansmith | cds: are any of your instances deleted=NULL? | 19:54 |
| *** arnaud____ has quit IRC | 19:55 | |
| cds | no | 19:55 |
| cds | the default for that column is null also | 19:55 |
| cds | which is why i was confused | 19:55 |
| dansmith | yeah, that's because $history I think | 19:55 |
| dansmith | cds: so of the four instances in there hat don't have extra records, does anything stand out? | 19:56 |
| dansmith | like, super old, super new? | 19:56 |
| dansmith | cds: are they in any state like RESIZED or something? | 19:56 |
| dansmith | cds: do they have a task_state ? | 19:56 |
| cds | let me poke around | 19:57 |
| *** dprince has quit IRC | 19:57 | |
| dansmith | cds: really appreciate your help with this | 19:57 |
| cds | my help? you are helping me :P | 19:57 |
| dansmith | well, | 19:58 |
| dansmith | this part helps you in the future, helps me now I guess :) | 19:59 |
| *** ijw_ has quit IRC | 19:59 | |
| cds | dansmith: instances in the error state do not appear in the instance_extra table | 20:02 |
| cds | that is the discrepancy | 20:02 |
| dansmith | cds: ah, okay, and I assume those have a task_state set, which is why | 20:03 |
| cds | are deleted instances info supposed to remain in instance_extra ? | 20:05 |
| melwitt | mriedem: I don't know if you've had a chance to revisit this log a warning for sec groups + port id https://review.openstack.org/#/c/154068/ | 20:06 |
| dansmith | cds: you mean if you have an instance with an extra record and then you delete it? | 20:07 |
| cds | yes | 20:07 |
| *** baoli has joined #openstack-nova | 20:08 | |
| dansmith | cds: yeah, we do a soft_delete() on that, which should just set deleted!=0 on the extra record | 20:08 |
| cds | last question... :) | 20:09 |
| cds | do you guys have any scripts laying around to purge an instance based on uuid | 20:10 |
| cds | from the nova db | 20:10 |
| mriedem | melwitt: i noticed there was some back and forth | 20:10 |
| *** eharney has joined #openstack-nova | 20:10 | |
| dansmith | cds: did you hear that? | 20:11 |
| dansmith | cds: that was the sound of everyone scattering from the room | 20:11 |
| cds | :( | 20:11 |
| dansmith | heh | 20:11 |
| dansmith | cds: unfortunately we don't, and we don't have any prescription for when and how to do that | 20:11 |
| dansmith | cds: we're also not proud of it. | 20:11 |
| cds | just want to get my db in a consistent state | 20:12 |
| alaski | I recently found http://openstack-in-production.blogspot.com/2015/05/purging-nova-databases-in-cell.html which could be used as a starting point | 20:12 |
| dansmith | alaski: ah, nice | 20:12 |
| dansmith | cds: it's actually a big problem and right now it's left as an exercise for the deployer, which sucks | 20:12 |
| dansmith | cds: a hint though, alaski runs a really big (singlehandedly, I hear) so you can press him for details :) | 20:13 |
| dansmith | er, "really big cloud" | 20:13 |
| alaski | well, I have two hands really | 20:13 |
| dansmith | hah | 20:13 |
| melwitt | mriedem: yeah, I was reading over it. I think it's okay, but can't say I'm sure how much logging a warning helps. maybe it's better than nothing | 20:14 |
| mriedem | melwitt: so.... | 20:14 |
| mriedem | i think i'm -1 on this | 20:14 |
| mriedem | well, i'm sure i'm -1 but for multiple reasons | 20:14 |
| *** mmedvede has quit IRC | 20:15 | |
| mriedem | melwitt: assuming we warn in L and error in M, we need this in the release notes so the commit message needs DocImpact/UpgradeImpact for that as a looming API failure | 20:15 |
| mriedem | but because it's suggesting we're going to break an API in M, i'm -1 on *that* | 20:15 |
| *** jwcroppe_ has joined #openstack-nova | 20:16 | |
| mriedem | b/c warts and all, i don't think we can do that | 20:16 |
| mriedem | until we have a new neutron api in nova where we can break backwards compat - so whatever beagles is working on | 20:16 |
| melwitt | mriedem: hm, yeah.. the back and forth made it get to this state | 20:17 |
| garyk | mriedem: is there any code from beagles? i have just seen the spec | 20:17 |
| *** jwcroppe has quit IRC | 20:18 | |
| beagles | garyk, no code yet - now that I have some approval on the specs, I can get moving | 20:18 |
| * beagles wishes he was going to summit, would've been great to discuss in person | 20:19 | |
| alaski | does anyone know a way to mock a method call invoked through super(<classname>, self).method()? | 20:19 |
| melwitt | cds: the blog post alaski links mentions about it, there is a nova-manage tool for archiving (deleting and moving to shadow tables) rows but it doesn't work on instances, it's broken | 20:20 |
| garyk | beagles: i guess by what you wrote means we will need to bach it out on irc | 20:20 |
| *** kmartin has quit IRC | 20:21 | |
| cds | yeah i rewriting that cern tool | 20:21 |
| cds | to just take the uuid | 20:21 |
| *** haypo has joined #openstack-nova | 20:21 | |
| *** ildikov has joined #openstack-nova | 20:21 | |
| *** romainh has left #openstack-nova | 20:24 | |
| *** claudiub has quit IRC | 20:24 | |
| *** jecarey has quit IRC | 20:24 | |
| openstackgerrit | Chris Friesen proposed openstack/nova: Fix race between resource audit and cpu pinning https://review.openstack.org/182766 | 20:25 |
| melwitt | alaski: do you have an example? do you mean the super class's method calls some other method and you want to mock that? afaik, I think you'd have to mock.patch.object a class, or use mock.patch with the path to the method | 20:25 |
| cfriesen | can anyone think of a better way to do this ^ given the current code? | 20:26 |
| *** pixelbeat has joined #openstack-nova | 20:26 | |
| *** dave-mccowan has joined #openstack-nova | 20:26 | |
| *** mmedvede has joined #openstack-nova | 20:27 | |
| alaski | melwitt: I'm trying to mock http://git.openstack.org/cgit/openstack/nova/tree/nova/compute/cells_api.py#n229 specifically | 20:27 |
| mriedem | alaski: i'm sorry | 20:28 |
| alaski | melwitt: but mocking it on the instantiated compute api didn't seem to work | 20:28 |
| mriedem | alaski: do you know what method_name is ahead of time? | 20:29 |
| dansmith | cfriesen: sorry, I was in the middle of that other thing earlier | 20:29 |
| cfriesen | dansmith: no worries | 20:29 |
| alaski | mriedem: heh. I'll probably just wrap it in a _get_parent_class method or something, because all I care about is that execution gets there. | 20:29 |
| alaski | mriedem: yes | 20:29 |
| melwitt | alaski: hm, I thought that would have worked | 20:29 |
| *** otter768 has joined #openstack-nova | 20:29 | |
| mriedem | and mock.patch.object(compute_api.API, method_name) doesn't work? | 20:30 |
| *** mwagner_lap has quit IRC | 20:30 | |
| alaski | I tried mocking on http://git.openstack.org/cgit/openstack/nova/tree/nova/tests/unit/compute/test_compute_cells.py#n106 | 20:31 |
| alaski | which is compute.API() | 20:31 |
| alaski | hmm, let me try the import | 20:32 |
| melwitt | alaski: it gets set to ComputeCellsAPI() on L118 though | 20:33 |
| alaski | melwitt: yeah, that threw me off at first, but I tried ORIG_COMPUTE_API as well | 20:33 |
| bauzas | alaski: oh, cells meeting in 30 mins right? | 20:33 |
| alaski | bauzas: yep | 20:33 |
| melwitt | alaski: oh, hm. | 20:33 |
| *** otter768 has quit IRC | 20:34 | |
| alaski | mriedem: that worked, you're my hero for the day | 20:35 |
| alaski | compute_api.API works, compute_api.API() does not | 20:35 |
| *** ijw has joined #openstack-nova | 20:36 | |
| mriedem | alaski: sweet | 20:36 |
| *** ijw has quit IRC | 20:37 | |
| mriedem | that makes two: | 20:37 |
| mriedem | (12:51:02 PM) stevemar: mriedem, my hero <3 | 20:37 |
| *** ijw has joined #openstack-nova | 20:37 | |
| alaski | heh | 20:37 |
| *** kmartin has joined #openstack-nova | 20:38 | |
| *** annegentle has quit IRC | 20:39 | |
| openstackgerrit | Andrew Laski proposed openstack/nova: Retry a cell delete if host constraint fails https://review.openstack.org/182772 | 20:39 |
| *** annegentle has joined #openstack-nova | 20:39 | |
| *** jecarey has joined #openstack-nova | 20:40 | |
| *** jwcroppe has joined #openstack-nova | 20:43 | |
| *** jwcroppe_ has quit IRC | 20:45 | |
| *** ajo has joined #openstack-nova | 20:47 | |
| *** achanda has quit IRC | 20:48 | |
| *** belmoreira has joined #openstack-nova | 20:49 | |
| *** vladikr has quit IRC | 20:49 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: detach cinder volume when instance destroyed https://review.openstack.org/169732 | 20:49 |
| openstackgerrit | Andrew Laski proposed openstack/nova: Retry a cell delete if host constraint fails https://review.openstack.org/182772 | 20:49 |
| openstackgerrit | garyk proposed openstack/nova: VMware: remove redundant check for block devices https://review.openstack.org/170641 | 20:49 |
| *** jwcroppe_ has joined #openstack-nova | 20:51 | |
| *** nkrinner has quit IRC | 20:51 | |
| *** ildikov has quit IRC | 20:52 | |
| *** jwcroppe has quit IRC | 20:53 | |
| *** pixelbeat has quit IRC | 20:54 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: remove invalid comment https://review.openstack.org/182839 | 20:56 |
| *** Sukhdev has joined #openstack-nova | 20:59 | |
| *** thangp has quit IRC | 21:00 | |
| *** eglynn_ has joined #openstack-nova | 21:01 | |
| *** Sukhdev_ has quit IRC | 21:01 | |
| *** eglynn has quit IRC | 21:01 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: VMware: remove invalid comment https://review.openstack.org/182839 | 21:01 |
| *** isq_ has joined #openstack-nova | 21:04 | |
| *** kashyap` has joined #openstack-nova | 21:05 | |
| *** dgonzalez_ has joined #openstack-nova | 21:05 | |
| *** aysyanne has joined #openstack-nova | 21:05 | |
| *** mriedem has quit IRC | 21:06 | |
| *** NobodyCa1 has joined #openstack-nova | 21:07 | |
| *** ildikov has joined #openstack-nova | 21:07 | |
| *** mishkin____ has joined #openstack-nova | 21:07 | |
| *** garyk has quit IRC | 21:09 | |
| *** rwmjones has quit IRC | 21:09 | |
| *** mleroy has quit IRC | 21:09 | |
| *** kashyap has quit IRC | 21:09 | |
| *** dgonzalez has quit IRC | 21:09 | |
| *** aysyd has quit IRC | 21:09 | |
| *** ekarlso has quit IRC | 21:09 | |
| *** mishkin___ has quit IRC | 21:09 | |
| *** larsks has quit IRC | 21:09 | |
| *** wolsen has quit IRC | 21:09 | |
| *** sgordon has quit IRC | 21:09 | |
| *** NobodyCam has quit IRC | 21:09 | |
| *** Kennan has quit IRC | 21:09 | |
| *** tanee has quit IRC | 21:09 | |
| *** greghaynes has quit IRC | 21:09 | |
| *** isq has quit IRC | 21:09 | |
| *** dgonzalez_ is now known as dgonzalez | 21:09 | |
| *** mleroy has joined #openstack-nova | 21:09 | |
| *** greghayn1 has joined #openstack-nova | 21:09 | |
| *** isq_ has quit IRC | 21:09 | |
| *** isq_ has joined #openstack-nova | 21:09 | |
| *** kashyap` has quit IRC | 21:09 | |
| *** kashyap` has joined #openstack-nova | 21:09 | |
| *** sgordon has joined #openstack-nova | 21:10 | |
| *** sgordon has quit IRC | 21:10 | |
| *** sgordon has joined #openstack-nova | 21:10 | |
| *** ekarlso has joined #openstack-nova | 21:10 | |
| *** wolsen has joined #openstack-nova | 21:10 | |
| *** larsks has joined #openstack-nova | 21:10 | |
| *** garyk has joined #openstack-nova | 21:10 | |
| *** Kennan has joined #openstack-nova | 21:10 | |
| *** tanee has joined #openstack-nova | 21:11 | |
| *** NobodyCa1 is now known as NobodyCam | 21:11 | |
| *** achanda has joined #openstack-nova | 21:11 | |
| *** Marga_ has quit IRC | 21:11 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: detach cinder volume when instance destroyed https://review.openstack.org/169732 | 21:11 |
| *** Marga_ has joined #openstack-nova | 21:12 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: remove redundant check for block devices https://review.openstack.org/170641 | 21:13 |
| *** pixelbeat has joined #openstack-nova | 21:16 | |
| *** vilobhmm11 has joined #openstack-nova | 21:20 | |
| *** gokrokve has quit IRC | 21:20 | |
| *** vilobhmm12 has joined #openstack-nova | 21:21 | |
| *** haomaiwa_ has quit IRC | 21:21 | |
| *** vilobhmm11 has quit IRC | 21:21 | |
| *** gokrokve has joined #openstack-nova | 21:21 | |
| openstackgerrit | Chris Friesen proposed openstack/nova: Fix race between resource audit and cpu pinning https://review.openstack.org/182766 | 21:22 |
| *** nickchase has quit IRC | 21:22 | |
| *** jwcroppe has joined #openstack-nova | 21:22 | |
| *** vilobhmm11 has joined #openstack-nova | 21:22 | |
| *** angdraug has joined #openstack-nova | 21:23 | |
| *** vilobhmm1 has quit IRC | 21:23 | |
| *** vilobhmm1 has joined #openstack-nova | 21:24 | |
| *** mriedem has joined #openstack-nova | 21:25 | |
| *** jwcroppe_ has quit IRC | 21:25 | |
| *** gokrokve has quit IRC | 21:25 | |
| *** vilobhmm12 has quit IRC | 21:26 | |
| melwitt | dansmith: in this patch, when you give the example of Instance(uuid=...) will it not go to save the uuid field as a changed field? asking because I have a similar thing in my cells patch doing an instance_update equivalent, and I obj_reset_changes(['uuid']) perhaps needlessly before the save() https://review.openstack.org/#/c/182766/2/nova/compute/resource_tracker.py,cm | 21:27 |
| *** vilobhmm11 has quit IRC | 21:27 | |
| dansmith | melwitt: it will try, yeah | 21:28 |
| dansmith | melwitt: which won't hurt anything, but it should be avoided | 21:28 |
| dansmith | obj.obj_reset_changes(fields['uuid']) | 21:28 |
| dansmith | will take care of it | 21:28 |
| dansmith | cfriesen: ^ | 21:28 |
| melwitt | okay | 21:29 |
| *** matrohon has joined #openstack-nova | 21:29 | |
| *** gokrokve has joined #openstack-nova | 21:29 | |
| *** greghayn1 is now known as greghaynes | 21:29 | |
| cfriesen | dansmith: before calling save(), you mean? | 21:30 |
| dansmith | cfriesen: yeah | 21:30 |
| cfriesen | dansmith: works. new rev coming up | 21:30 |
| dansmith | cfriesen: you can also just reset all changes before setting the numa | 21:31 |
| dansmith | either way | 21:31 |
| dansmith | cfriesen: wait | 21:31 |
| cfriesen | yes | 21:31 |
| *** annegentle has quit IRC | 21:32 | |
| dansmith | cfriesen: comment on your comment | 21:32 |
| dansmith | cfriesen: be more forceful and carry more hate in your voice, please | 21:32 |
| cfriesen | :) | 21:33 |
| cfriesen | dansmith: to clarify...how would RT using instance objects throughout change things? instance_ref is already an object | 21:33 |
| dansmith | it's not | 21:33 |
| * dansmith goes to back up that bold statement | 21:34 | |
| cfriesen | sure it is. :) | 21:34 |
| *** kmartin has quit IRC | 21:34 | |
| dansmith | oh, this is passed in from compute just for this call isn't it? | 21:35 |
| cfriesen | yep | 21:35 |
| dansmith | then I'm -2 on this :) | 21:35 |
| cfriesen | heh | 21:35 |
| dansmith | just convert that whole thing to do a .save | 21:35 |
| dansmith | dump the instance_update | 21:35 |
| cfriesen | okay, but that'll save all the changes, no? | 21:35 |
| cfriesen | I wasn't sure it was okay to save everything at this point | 21:36 |
| dansmith | it will, which will be the three there plus your numa topology | 21:36 |
| dansmith | cfriesen: the numa topology isn't set on instance_ref until after that call | 21:37 |
| dansmith | oh, you moved this, right? | 21:38 |
| dansmith | it was in _set_instance_host_and_node() before right? | 21:38 |
| cfriesen | dansmith: yeah, I moved it to right before where we drop the semaphore | 21:39 |
| dansmith | okay | 21:39 |
| dansmith | so, I think what you should do is set the instance.numa_topology before the call to _set_instance_host_and_node() and then just do instance.save() in there | 21:39 |
| dansmith | maybe with a comment by the call that says it'll be saved when we do the host and node set | 21:40 |
| *** ildikov has quit IRC | 21:40 | |
| dansmith | that makes the whole thing a single save call | 21:40 |
| mriedem | alaski: https://review.openstack.org/#/c/169517/ should probably be on stable/kilo right? | 21:40 |
| dansmith | cfriesen: I don't think there's any reason not to save the whole thing there | 21:41 |
| mriedem | alaski: nevermind | 21:41 |
| alaski | mriedem: yes | 21:41 |
| mriedem | was merged in kilo | 21:41 |
| alaski | oh, cool | 21:41 |
| mriedem | when you said 'just fixed that' in the other review, i thought that was more literal | 21:41 |
| alaski | time is relative :) but it didn't seem like 5 weeks ago | 21:42 |
| cfriesen | dansmith: okay, can give that a try. | 21:42 |
| dansmith | cfriesen: it would be awesome if you followed this fix up with converting any places in that call chain that are always an object to use attribute syntax | 21:43 |
| dansmith | cfriesen: it's super confusing to see that be used as a dict there | 21:43 |
| cfriesen | dansmith: yeah...legacy code I'm assuming | 21:43 |
| dansmith | yes | 21:43 |
| dansmith | and it used to have to stay that way because of where instance was coming from, but it looks like maybe more of this can be cleaned up than I thought | 21:43 |
| dansmith | so at least right here would be good | 21:44 |
| * dansmith runs off for a bit | 21:44 | |
| *** Sukhdev_ has joined #openstack-nova | 21:44 | |
| *** gokrokve has quit IRC | 21:45 | |
| *** gokrokve has joined #openstack-nova | 21:45 | |
| *** Sukhdev has quit IRC | 21:46 | |
| *** gokrokve has quit IRC | 21:50 | |
| *** dave-mccowan has quit IRC | 21:53 | |
| *** ildikov has joined #openstack-nova | 21:53 | |
| *** belmoreira has quit IRC | 21:54 | |
| *** IanGovett has quit IRC | 21:57 | |
| *** IanGovett has joined #openstack-nova | 21:57 | |
| jogo | have we documented v2.0 as deprecated yet? | 21:57 |
| *** ajo has quit IRC | 21:58 | |
| *** mriedem is now known as mriedem_away | 21:58 | |
| *** aysyanne has quit IRC | 22:03 | |
| *** lucas-dinner has quit IRC | 22:03 | |
| *** nihilifer has joined #openstack-nova | 22:04 | |
| *** ddutta has joined #openstack-nova | 22:05 | |
| *** igordcard has joined #openstack-nova | 22:10 | |
| *** Marga_ has quit IRC | 22:11 | |
| *** gokrokve has joined #openstack-nova | 22:11 | |
| *** Marga_ has joined #openstack-nova | 22:11 | |
| *** lpetrut has quit IRC | 22:13 | |
| *** achanda has quit IRC | 22:13 | |
| *** Nic1 has quit IRC | 22:14 | |
| *** eharney has quit IRC | 22:14 | |
| *** salv-orlando has joined #openstack-nova | 22:15 | |
| *** salv-orlando has quit IRC | 22:15 | |
| *** jecarey has quit IRC | 22:16 | |
| *** eglynn_ has quit IRC | 22:16 | |
| *** vilobhmm1 has quit IRC | 22:16 | |
| *** zzzeek has quit IRC | 22:17 | |
| openstackgerrit | Chris Friesen proposed openstack/nova: Fix race between resource audit and cpu pinning https://review.openstack.org/182766 | 22:17 |
| *** vilobhmm1 has joined #openstack-nova | 22:17 | |
| *** matrohon has quit IRC | 22:17 | |
| openstackgerrit | Chris Friesen proposed openstack/nova: Fix race between resource audit and cpu pinning https://review.openstack.org/182766 | 22:18 |
| *** burt has quit IRC | 22:19 | |
| dansmith | cfriesen: https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L175 | 22:19 |
| dansmith | might be what I was thinking of | 22:19 |
| *** jamielennox|away is now known as jamielennox | 22:19 | |
| cfriesen | dansmith: likely. the most recent version should be what you were asking for...I haven't had a chance to test it out critically yet. | 22:21 |
| dansmith | cfriesen: yep, just looked | 22:21 |
| dansmith | cfriesen: needs tests and then I think it's good | 22:21 |
| cfriesen | dansmith: you looking for additional unit tests to cover the race case? | 22:23 |
| dansmith | cfriesen: yeah | 22:23 |
| dansmith | cfriesen: probably also good to verify that the numa bit gets caught in that save too, | 22:23 |
| dansmith | so that someone doesn't reorder it ignoring your comment | 22:23 |
| cfriesen | dansmith: makes sense. going to be a while, got internal stuff to close off before summit | 22:24 |
| dansmith | okay | 22:24 |
| cfriesen | dansmith: thanks for the assist | 22:24 |
| dansmith | np, thanks for not listening to my first assessment of the situation :) | 22:25 |
| *** nelsnelson has quit IRC | 22:25 | |
| cfriesen | lol | 22:26 |
| *** baoli has quit IRC | 22:26 | |
| *** otter768 has joined #openstack-nova | 22:30 | |
| *** pixelbeat has quit IRC | 22:30 | |
| *** Sukhdev has joined #openstack-nova | 22:33 | |
| *** patrickeast has quit IRC | 22:33 | |
| *** patrickeast_ has joined #openstack-nova | 22:33 | |
| *** ekarlso has quit IRC | 22:34 | |
| *** ekarlso has joined #openstack-nova | 22:34 | |
| *** otter768 has quit IRC | 22:35 | |
| *** cfriesen has quit IRC | 22:35 | |
| *** cfriesen has joined #openstack-nova | 22:35 | |
| *** Sukhdev_ has quit IRC | 22:37 | |
| *** dims_ has joined #openstack-nova | 22:40 | |
| *** achanda has joined #openstack-nova | 22:40 | |
| -openstackstatus- NOTICE: Gerrit and Zuul are going offline for reboots to fix a security vulnerability. | 22:40 | |
| *** ChanServ changes topic to "Gerrit and Zuul are going offline for reboots to fix a security vulnerability." | 22:40 | |
| *** dims has quit IRC | 22:41 | |
| *** Sukhdev has quit IRC | 22:43 | |
| *** dims has joined #openstack-nova | 22:43 | |
| *** claudiub has joined #openstack-nova | 22:45 | |
| *** dims_ has quit IRC | 22:47 | |
| *** openstackgerrit has quit IRC | 22:47 | |
| *** ddutta has quit IRC | 22:48 | |
| *** openstackgerrit has joined #openstack-nova | 22:49 | |
| *** claudiub has quit IRC | 22:49 | |
| *** claudiub has joined #openstack-nova | 22:50 | |
| *** pixelbeat has joined #openstack-nova | 22:51 | |
| *** Raildo_ has joined #openstack-nova | 22:53 | |
| *** tonytan4ever has quit IRC | 22:54 | |
| *** obedmr has quit IRC | 22:54 | |
| *** ericksonfgds has joined #openstack-nova | 22:58 | |
| *** ChanServ changes topic to "This channel is for Nova development. For support of Nova deployments, please use #openstack. -=- master is open for Lemming development" | 22:58 | |
| -openstackstatus- NOTICE: Gerrit and Zuul are back online. | 22:58 | |
| *** patrickeast_ has quit IRC | 22:58 | |
| *** patrickeast has joined #openstack-nova | 23:01 | |
| *** tjones2 has joined #openstack-nova | 23:07 | |
| *** tjones2 has left #openstack-nova | 23:08 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Fix loading things in instance_extra for old instances https://review.openstack.org/182787 | 23:08 |
| *** achanda has quit IRC | 23:09 | |
| *** tjones1 has quit IRC | 23:09 | |
| *** gokrokve has quit IRC | 23:13 | |
| *** gokrokve has joined #openstack-nova | 23:14 | |
| *** gokrokve has quit IRC | 23:15 | |
| *** pixelbeat has quit IRC | 23:16 | |
| *** signed8bit is now known as signed8bit_ZZZzz | 23:16 | |
| *** vipuls is now known as vipul | 23:20 | |
| *** Marga_ has quit IRC | 23:20 | |
| *** Marga_ has joined #openstack-nova | 23:21 | |
| openstackgerrit | Joe Gordon proposed openstack/nova: Add note to doc explaining scope https://review.openstack.org/182881 | 23:22 |
| openstackgerrit | Joe Gordon proposed openstack/nova: Consolidate API sections in docs https://review.openstack.org/182882 | 23:22 |
| openstackgerrit | Joe Gordon proposed openstack/nova: Use a depth of 2 in devref https://review.openstack.org/182883 | 23:22 |
| openstackgerrit | Joe Gordon proposed openstack/nova: Prune development.environment doc https://review.openstack.org/182884 | 23:22 |
| openstackgerrit | Joe Gordon proposed openstack/nova: Add API section to devref https://review.openstack.org/182885 | 23:22 |
| openstackgerrit | Victor Stinner proposed openstack/nova: Replace dict(obj.iteritems()) with dict(obj) https://review.openstack.org/182887 | 23:24 |
| *** achanda has joined #openstack-nova | 23:26 | |
| *** igordcard has quit IRC | 23:29 | |
| *** signed8bit_ZZZzz is now known as signed8bit | 23:31 | |
| dansmith | cfriesen: | 23:33 |
| openstackgerrit | Dan Smith proposed openstack/nova: Fix race between resource audit and cpu pinning https://review.openstack.org/182766 | 23:33 |
| *** JayF has quit IRC | 23:33 | |
| dansmith | += tests | 23:33 |
| *** BadCub is now known as BadCub-Hiding | 23:38 | |
| *** Raildo_ has quit IRC | 23:44 | |
| *** rfolco has quit IRC | 23:45 | |
| *** ociuhandu has quit IRC | 23:45 | |
| *** dave-mccowan has joined #openstack-nova | 23:47 | |
| *** tsekiyam_ has joined #openstack-nova | 23:53 | |
| *** exploreshaifali has joined #openstack-nova | 23:54 | |
| jlvillal | dansmith: Any chance you can "restore" https://review.openstack.org/#/c/130877/ ? | 23:54 |
| jlvillal | dansmith: I am going to start working on it. And the original owner is on vacation, until the summit starts. | 23:54 |
| dansmith | done | 23:55 |
| jlvillal | dansmith: Thanks! | 23:55 |
| dansmith | np | 23:55 |
| *** ericksonfgds has quit IRC | 23:55 | |
| *** tsekiyama has quit IRC | 23:57 | |
| openstackgerrit | Victor Stinner proposed openstack/nova: Replace dict.iteritems() with six.iteritems(dict) https://review.openstack.org/179784 | 23:57 |
| openstackgerrit | Victor Stinner proposed openstack/nova: Replace dict(obj.iteritems()) with dict(obj) https://review.openstack.org/182887 | 23:57 |
| *** mtanino has quit IRC | 23:57 | |
| *** tsekiyam_ has quit IRC | 23:58 | |
| *** tjones1 has joined #openstack-nova | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!