*** tsekiyama has quit IRC | 00:01 | |
*** Marga_ has joined #openstack-nova | 00:02 | |
*** DCWillia_ has quit IRC | 00:02 | |
*** annegentle has joined #openstack-nova | 00:03 | |
*** annegentle has quit IRC | 00:08 | |
*** tjones1 has quit IRC | 00:09 | |
*** signed8bit_ZZZzz is now known as signed8bit | 00:12 | |
*** signed8bit is now known as signed8bit_ZZZzz | 00:12 | |
*** dboik has joined #openstack-nova | 00:14 | |
*** Marga_ has quit IRC | 00:17 | |
*** Marga_ has joined #openstack-nova | 00:17 | |
*** dboik has quit IRC | 00:18 | |
*** tjones1 has joined #openstack-nova | 00:18 | |
*** oro has quit IRC | 00:20 | |
*** VW_ has quit IRC | 00:25 | |
*** yingjun has joined #openstack-nova | 00:27 | |
*** igordcard has quit IRC | 00:28 | |
*** zhenguo has joined #openstack-nova | 00:28 | |
*** tjones1 has quit IRC | 00:32 | |
*** promulo__ has joined #openstack-nova | 00:32 | |
*** tjones1 has joined #openstack-nova | 00:35 | |
*** promulo_ has quit IRC | 00:36 | |
*** annegentle has joined #openstack-nova | 00:39 | |
*** tjones1 has quit IRC | 00:40 | |
*** DCWilliams_VA has joined #openstack-nova | 00:40 | |
*** liusheng has quit IRC | 00:41 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support https://review.openstack.org/180695 | 00:41 |
---|---|---|
*** the_hugger has quit IRC | 00:41 | |
*** liusheng has joined #openstack-nova | 00:41 | |
*** VW_ has joined #openstack-nova | 00:44 | |
openstackgerrit | Paul Murray proposed openstack/nova: Use compute_node consistently in ResourceTracker https://review.openstack.org/159765 | 00:47 |
*** IanGovett1 has joined #openstack-nova | 00:49 | |
*** IanGovett has quit IRC | 00:49 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support https://review.openstack.org/180695 | 00:49 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 00:51 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 00:51 |
*** annegentle has quit IRC | 00:56 | |
*** zsais has quit IRC | 01:00 | |
*** shakamunyi has quit IRC | 01:02 | |
*** dboik has joined #openstack-nova | 01:05 | |
*** signed8bit_ZZZzz is now known as signed8bit | 01:06 | |
*** dboik_ has joined #openstack-nova | 01:06 | |
*** signed8bit is now known as signed8bit_ZZZzz | 01:06 | |
*** dboik has quit IRC | 01:09 | |
*** DCWilliams_VA has quit IRC | 01:12 | |
*** claudiub has quit IRC | 01:12 | |
*** asettle has joined #openstack-nova | 01:12 | |
*** VW_ has quit IRC | 01:16 | |
*** park_heijlong has quit IRC | 01:16 | |
*** angdraug has quit IRC | 01:18 | |
*** otter768 has joined #openstack-nova | 01:26 | |
*** achanda has quit IRC | 01:29 | |
*** otter768_ has joined #openstack-nova | 01:29 | |
*** otter768 has quit IRC | 01:30 | |
*** popw has joined #openstack-nova | 01:32 | |
*** FL1SK has quit IRC | 01:41 | |
*** Marga_ has quit IRC | 01:42 | |
*** Marga_ has joined #openstack-nova | 01:43 | |
openstackgerrit | Eli Qiao proposed openstack/nova-specs: Validate the service state before deleting it https://review.openstack.org/163274 | 01:45 |
*** tjones1 has joined #openstack-nova | 01:48 | |
*** zzzeek has quit IRC | 01:53 | |
*** otter768_ has quit IRC | 01:56 | |
openstackgerrit | melanie witt proposed openstack/nova: Send Instance object to cells instance_update_at_top https://review.openstack.org/176518 | 01:56 |
*** vilobhmm1 has quit IRC | 01:56 | |
*** annegentle has joined #openstack-nova | 01:56 | |
*** patrickeast has quit IRC | 01:59 | |
openstackgerrit | Eli Qiao proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_get_all_by_name https://review.openstack.org/160205 | 01:59 |
*** zhangtralon has joined #openstack-nova | 02:01 | |
*** dboik_ has quit IRC | 02:01 | |
*** erlon has quit IRC | 02:01 | |
*** dboik has joined #openstack-nova | 02:01 | |
openstackgerrit | Eli Qiao proposed openstack/nova: API: remove unuseful expected error code from v2.1 service delete api. https://review.openstack.org/180827 | 02:02 |
*** annegentle has quit IRC | 02:02 | |
*** salv-orlando has quit IRC | 02:03 | |
*** neelashah has joined #openstack-nova | 02:04 | |
*** ddaskal has joined #openstack-nova | 02:05 | |
*** unicell has quit IRC | 02:07 | |
*** browne1 has quit IRC | 02:07 | |
*** shakamunyi has joined #openstack-nova | 02:09 | |
*** mriedem has quit IRC | 02:14 | |
*** Kevin_Zheng has joined #openstack-nova | 02:15 | |
*** yamahata has joined #openstack-nova | 02:17 | |
*** yamahata has quit IRC | 02:20 | |
*** yamahata has joined #openstack-nova | 02:20 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/180601 | 02:20 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 02:21 |
*** yingjun has quit IRC | 02:26 | |
*** baoli has joined #openstack-nova | 02:27 | |
*** achanda has joined #openstack-nova | 02:29 | |
*** yingjun has joined #openstack-nova | 02:30 | |
*** baoli_ has joined #openstack-nova | 02:30 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 02:31 |
*** mtanino has quit IRC | 02:32 | |
*** baoli has quit IRC | 02:33 | |
*** VW_ has joined #openstack-nova | 02:35 | |
*** bkopilov has quit IRC | 02:35 | |
*** jogo has quit IRC | 02:45 | |
*** Marga_ has quit IRC | 02:45 | |
*** achanda has quit IRC | 02:46 | |
*** achanda has joined #openstack-nova | 02:48 | |
*** r-daneel has quit IRC | 02:49 | |
openstackgerrit | wanghao proposed openstack/nova: Allow bdm boot_index=1 when creating vm from image https://review.openstack.org/171984 | 02:49 |
openstackgerrit | Eli Qiao proposed openstack/nova: API: remove unuseful expected error code from v2.1 service delete api. https://review.openstack.org/180827 | 02:50 |
*** zzzeek has joined #openstack-nova | 02:50 | |
openstackgerrit | Eli Qiao proposed openstack/nova: API: remove unuseful expected error code from v2.1 service delete api. https://review.openstack.org/180827 | 02:50 |
*** dims_ has quit IRC | 02:51 | |
*** dims has joined #openstack-nova | 02:52 | |
*** annegentle has joined #openstack-nova | 02:57 | |
*** vishwanathj has joined #openstack-nova | 03:02 | |
*** jogo has joined #openstack-nova | 03:02 | |
*** markvoelker has quit IRC | 03:04 | |
*** zzzeek has quit IRC | 03:05 | |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_get_all_* https://review.openstack.org/160202 | 03:05 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_create/update https://review.openstack.org/160203 | 03:05 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_destroy_all_* https://review.openstack.org/160201 | 03:05 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_usage_update https://review.openstack.org/160215 | 03:05 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 03:09 |
*** sdake_ has joined #openstack-nova | 03:09 | |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_get_all_* https://review.openstack.org/160202 | 03:11 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_create/update https://review.openstack.org/160203 | 03:11 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_usage_update https://review.openstack.org/160215 | 03:11 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 03:11 |
*** dims has quit IRC | 03:11 | |
*** sdake has quit IRC | 03:13 | |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_create/update https://review.openstack.org/160206 | 03:17 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_get_all_by_name https://review.openstack.org/160205 | 03:17 |
openstackgerrit | Alex Xu proposed openstack/nova: Cleanup quota_class unittest with appropriate request context https://review.openstack.org/150351 | 03:17 |
*** dsanders has quit IRC | 03:18 | |
*** jyuso1 has quit IRC | 03:23 | |
*** annegentle has quit IRC | 03:25 | |
*** Murali has joined #openstack-nova | 03:25 | |
*** otter768 has joined #openstack-nova | 03:26 | |
openstackgerrit | Alex Xu proposed openstack/nova: API: remove unuseful expected error code from v2.1 service delete api https://review.openstack.org/180827 | 03:28 |
*** achanda has quit IRC | 03:28 | |
*** tjones1 has left #openstack-nova | 03:28 | |
*** neelashah has quit IRC | 03:29 | |
*** baoli_ has quit IRC | 03:30 | |
*** asettle has quit IRC | 03:34 | |
openstackgerrit | Vincent Hou proposed openstack/nova: Add a test case to ensure the right parameters for swap_volume https://review.openstack.org/174307 | 03:34 |
*** sdake has joined #openstack-nova | 03:36 | |
*** browne has joined #openstack-nova | 03:36 | |
*** sdake_ has quit IRC | 03:39 | |
*** IanGovett1 has quit IRC | 03:40 | |
*** mmedvede_ has quit IRC | 03:41 | |
*** lpetrut has joined #openstack-nova | 03:41 | |
*** dikonoor has joined #openstack-nova | 03:42 | |
*** marun has quit IRC | 03:44 | |
*** links has joined #openstack-nova | 03:45 | |
*** FL1SK has joined #openstack-nova | 03:46 | |
*** mmedvede_ has joined #openstack-nova | 03:48 | |
*** alanf-mc has quit IRC | 03:49 | |
*** yingjun has quit IRC | 03:50 | |
*** deepthi has joined #openstack-nova | 03:52 | |
*** shyama has joined #openstack-nova | 04:00 | |
*** bkopilov has joined #openstack-nova | 04:01 | |
*** otter768 has quit IRC | 04:01 | |
*** emagana has quit IRC | 04:03 | |
*** markvoelker has joined #openstack-nova | 04:04 | |
*** achanda has joined #openstack-nova | 04:05 | |
*** READ10 has quit IRC | 04:07 | |
*** armax has joined #openstack-nova | 04:07 | |
*** markvoelker has quit IRC | 04:09 | |
*** tojuvone has joined #openstack-nova | 04:13 | |
*** armax has quit IRC | 04:16 | |
*** jogo has quit IRC | 04:16 | |
*** tjones1 has joined #openstack-nova | 04:20 | |
openstackgerrit | Rui Chen proposed openstack/nova: Clean up ComputeManager._get_instance_nw_info https://review.openstack.org/173680 | 04:21 |
*** tjones1 has left #openstack-nova | 04:21 | |
*** alanf-mc has joined #openstack-nova | 04:23 | |
*** dave-mccowan has quit IRC | 04:24 | |
*** garyk has joined #openstack-nova | 04:27 | |
*** alanf-mc has quit IRC | 04:30 | |
*** alanf-mc has joined #openstack-nova | 04:30 | |
*** jogo has joined #openstack-nova | 04:32 | |
*** oomichi has joined #openstack-nova | 04:33 | |
*** shyama has quit IRC | 04:36 | |
*** lpetrut has quit IRC | 04:40 | |
*** arnaud____ has joined #openstack-nova | 04:40 | |
*** emagana has joined #openstack-nova | 04:41 | |
*** jog0 has joined #openstack-nova | 04:44 | |
*** jogo has quit IRC | 04:45 | |
*** emagana has quit IRC | 04:46 | |
*** shyama has joined #openstack-nova | 04:47 | |
*** ildikov has quit IRC | 04:49 | |
*** kiran-r has joined #openstack-nova | 04:51 | |
*** Murali has quit IRC | 04:54 | |
*** yingjun has joined #openstack-nova | 04:59 | |
*** andrewbogott_afk is now known as andrewbogott | 05:00 | |
*** andrewbogott is now known as andrewbogott_afk | 05:00 | |
*** mmedvede has quit IRC | 05:01 | |
*** ildikov has joined #openstack-nova | 05:02 | |
*** coolsvap|afk is now known as coolsvap | 05:03 | |
*** nkrinner has joined #openstack-nova | 05:04 | |
*** yingjun has quit IRC | 05:04 | |
*** markvoelker has joined #openstack-nova | 05:05 | |
*** yingjun has joined #openstack-nova | 05:05 | |
*** unicell has joined #openstack-nova | 05:06 | |
*** andrewbogott_afk is now known as andrewbogott | 05:09 | |
*** ildikov has quit IRC | 05:09 | |
*** zhangtralon has quit IRC | 05:09 | |
*** arnaud____ has quit IRC | 05:10 | |
*** markvoelker has quit IRC | 05:10 | |
*** zhangtralon has joined #openstack-nova | 05:11 | |
*** vilobhmm1 has joined #openstack-nova | 05:12 | |
*** yingjun has quit IRC | 05:13 | |
*** yingjun has joined #openstack-nova | 05:13 | |
*** dsanders has joined #openstack-nova | 05:15 | |
*** zhangtralon has quit IRC | 05:17 | |
*** andrewbogott is now known as andrewbogott_afk | 05:18 | |
*** yingjun has quit IRC | 05:18 | |
*** Nikolay_St has quit IRC | 05:19 | |
*** sdake has quit IRC | 05:20 | |
*** sdake has joined #openstack-nova | 05:24 | |
openstackgerrit | garyk proposed openstack/nova: Object: add volume_type_id to block_device objects https://review.openstack.org/139994 | 05:24 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_get_all_* https://review.openstack.org/160202 | 05:24 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_create/update https://review.openstack.org/160203 | 05:24 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_destroy_all_* https://review.openstack.org/160201 | 05:24 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_usage_update https://review.openstack.org/160215 | 05:24 |
*** sdake_ has joined #openstack-nova | 05:26 | |
*** mdbooth has quit IRC | 05:27 | |
*** sdake has quit IRC | 05:29 | |
*** mdbooth has joined #openstack-nova | 05:31 | |
*** Longgeek has joined #openstack-nova | 05:33 | |
openstackgerrit | Gábor Takács proposed openstack/nova: Refactor os-services unittest https://review.openstack.org/180869 | 05:34 |
*** emagana has joined #openstack-nova | 05:35 | |
openstackgerrit | garyk proposed openstack/nova: Compute: remove reverts_task_state from interface attach/detach https://review.openstack.org/180537 | 05:37 |
*** moshele has joined #openstack-nova | 05:38 | |
*** emagana has quit IRC | 05:40 | |
*** arnaud____ has joined #openstack-nova | 05:42 | |
*** jog0 has quit IRC | 05:42 | |
*** _gryf|afk is now known as _gryf | 05:43 | |
*** jogo has joined #openstack-nova | 05:50 | |
*** jogo has quit IRC | 05:50 | |
*** jogo has joined #openstack-nova | 05:50 | |
*** salv-orlando has joined #openstack-nova | 05:50 | |
*** markvoelker has joined #openstack-nova | 05:51 | |
*** VW_ has quit IRC | 05:53 | |
*** mmedvede_ has quit IRC | 05:55 | |
*** markvoelker has quit IRC | 05:56 | |
*** mpsairam has joined #openstack-nova | 05:56 | |
*** shyama has quit IRC | 05:56 | |
*** moshele has quit IRC | 05:56 | |
*** shyama has joined #openstack-nova | 05:56 | |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_create/update https://review.openstack.org/160206 | 05:57 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_get_all_by_name https://review.openstack.org/160205 | 05:57 |
openstackgerrit | Alex Xu proposed openstack/nova: Cleanup quota_class unittest with appropriate request context https://review.openstack.org/150351 | 05:57 |
*** yingjun has joined #openstack-nova | 05:57 | |
*** baoli has joined #openstack-nova | 05:57 | |
*** vilobhmm1 has quit IRC | 05:59 | |
alex_xu | oomichi: hi, I put something in the summit etherpad https://etherpad.openstack.org/p/liberty-nova-api hope you can check it if there is anything isn't worth discuss when you have time, thanks in advance! | 06:01 |
*** baoli has quit IRC | 06:02 | |
*** otter768 has joined #openstack-nova | 06:02 | |
*** Kennan has quit IRC | 06:02 | |
*** Kennan has joined #openstack-nova | 06:03 | |
oomichi | alex_xu: thanks, many items :-) | 06:03 |
*** jyuso1 has joined #openstack-nova | 06:03 | |
*** dsanders has quit IRC | 06:04 | |
alex_xu | oomichi: yea, many items, just feel free to edit if anything wrong, let me know whether I work with right way :) | 06:04 |
oomichi | alex_xu: yeah, nice work. I am checking it now | 06:05 |
alex_xu | oomichi: thanks, no hurry | 06:05 |
oomichi | alex_xu: btw we will have api meeting tomorrow? | 06:05 |
alex_xu | oomichi: do you want to have one, as I know gilliard_afk is on vacation | 06:06 |
*** Longgeek has quit IRC | 06:06 | |
alex_xu | oomichi: maybe we can have one if you have anything want to suggestion for the etherpad | 06:06 |
*** otter768 has quit IRC | 06:07 | |
oomichi | alex_xu: unfortunately I cannot be sure that I can attend to it tomorrow. so I will try suggesting on each review | 06:07 |
*** pkoniszewski has joined #openstack-nova | 06:07 | |
alex_xu | oomichi: ok, it's fine, no problem, let me collect the suggestion | 06:08 |
*** sdake_ is now known as sdake | 06:10 | |
eliqiao1 | hi oomichi: thanks the reviewing , it's re https://review.openstack.org/#/c/175744/12/nova/tests/unit/api/openstack/compute/contrib/test_hypervisors.py | 06:12 |
oomichi | eliqiao1: hi | 06:12 |
openstackgerrit | Rui Chen proposed openstack/nova: Fix empty instance nw_info cache https://review.openstack.org/146036 | 06:13 |
eliqiao1 | oomichi: I am afraid we can not move self.policy.set_rules to setup() because test case will be reset after being executed. | 06:13 |
*** yingjun has quit IRC | 06:13 | |
oomichi | eliqiao1: thanks for the notice, I'd like to confirm it on my test env. | 06:14 |
*** mmedvede has joined #openstack-nova | 06:15 | |
eliqiao1 | oomichi: sure, thanks | 06:15 |
*** ddaskal has quit IRC | 06:18 | |
*** yingjun has joined #openstack-nova | 06:19 | |
openstackgerrit | Eli Qiao proposed openstack/nova: Remove db layer hard-code permission checks for floating_ip_dns https://review.openstack.org/150710 | 06:21 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Imported Translations from Transifex https://review.openstack.org/179638 | 06:21 |
*** ildikov has joined #openstack-nova | 06:24 | |
*** Murali has joined #openstack-nova | 06:28 | |
*** jcoufal has joined #openstack-nova | 06:28 | |
*** zhangtralon has joined #openstack-nova | 06:29 | |
*** exploreshaifali has joined #openstack-nova | 06:29 | |
oomichi | eliqiao1: sorry, that is my bad. LGTM now | 06:30 |
*** oro has joined #openstack-nova | 06:31 | |
*** oomichi has quit IRC | 06:31 | |
*** ildikov has quit IRC | 06:32 | |
*** emagana has joined #openstack-nova | 06:37 | |
*** ankit_ag has joined #openstack-nova | 06:38 | |
*** exploreshaifali has quit IRC | 06:40 | |
*** exploreshaifali has joined #openstack-nova | 06:40 | |
ankit_ag | alaski: hi | 06:42 |
*** emagana has quit IRC | 06:42 | |
*** arnaud____ has quit IRC | 06:43 | |
alex_xu | ankit_ag: US didn't wake up yet, need waiting for more few hours :) | 06:45 |
ankit_ag | alex_xu: yeah :) | 06:46 |
*** romainh has joined #openstack-nova | 06:46 | |
*** lpetrut has joined #openstack-nova | 06:47 | |
*** moshele has joined #openstack-nova | 06:47 | |
*** ildikov has joined #openstack-nova | 06:48 | |
*** markvoelker has joined #openstack-nova | 06:52 | |
*** shyama has quit IRC | 06:55 | |
*** Mike_D_laptop has joined #openstack-nova | 06:56 | |
*** markvoelker has quit IRC | 06:56 | |
*** tojuvone has quit IRC | 06:59 | |
*** sahid has joined #openstack-nova | 07:00 | |
*** mpavone has joined #openstack-nova | 07:02 | |
*** matrohon has joined #openstack-nova | 07:03 | |
*** Longgeek has joined #openstack-nova | 07:07 | |
*** Mike_D_laptop has quit IRC | 07:10 | |
*** xgsa has quit IRC | 07:18 | |
*** exploreshaifali has quit IRC | 07:18 | |
*** jichen has joined #openstack-nova | 07:20 | |
*** Longgeek has quit IRC | 07:21 | |
*** Longgeek has joined #openstack-nova | 07:21 | |
*** dulek has joined #openstack-nova | 07:24 | |
*** markus_z has joined #openstack-nova | 07:31 | |
*** claudiub has joined #openstack-nova | 07:31 | |
*** lpetrut has quit IRC | 07:31 | |
*** emagana has joined #openstack-nova | 07:32 | |
*** Nikolay_St has joined #openstack-nova | 07:32 | |
*** cfriesen has quit IRC | 07:35 | |
*** emagana has quit IRC | 07:36 | |
*** xgsa has joined #openstack-nova | 07:37 | |
*** takedakn has joined #openstack-nova | 07:37 | |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_create/update https://review.openstack.org/160206 | 07:38 |
openstackgerrit | Alex Xu proposed openstack/nova: Cleanup quota_class unittest with appropriate request context https://review.openstack.org/150351 | 07:38 |
openstackgerrit | Ankit Agrawal proposed openstack/nova: Detach volume after deleting instance with no host https://review.openstack.org/145738 | 07:39 |
*** VW_ has joined #openstack-nova | 07:42 | |
openstackgerrit | sahid proposed openstack/nova: libvirt: add method to get cpu stats to host https://review.openstack.org/177223 | 07:42 |
openstackgerrit | sahid proposed openstack/nova: monitor: remove dependance with libvirt https://review.openstack.org/178154 | 07:42 |
openstackgerrit | sahid proposed openstack/nova: libvirt: add method to compare cpu to Host https://review.openstack.org/177331 | 07:42 |
openstackgerrit | sahid proposed openstack/nova: libvirt: add method to list pci devices to Host https://review.openstack.org/177330 | 07:42 |
openstackgerrit | sahid proposed openstack/nova: libvirt: add method to get device by name to Host https://review.openstack.org/177329 | 07:42 |
openstackgerrit | sahid proposed openstack/nova: libvirt: add method to define instance to host https://review.openstack.org/174314 | 07:42 |
openstackgerrit | sahid proposed openstack/nova: libvirt: introduces a Guest to wrap arround virConnect https://review.openstack.org/174313 | 07:42 |
openstackgerrit | sahid proposed openstack/nova: libvirt: add todo note to avoid call to libvirt from the driver https://review.openstack.org/177332 | 07:42 |
*** lpetrut has joined #openstack-nova | 07:42 | |
*** oro has quit IRC | 07:42 | |
*** takedakn has quit IRC | 07:44 | |
*** mmedvede has quit IRC | 07:45 | |
tdurakov | bauzas, hi) another one for you:) https://bugs.launchpad.net/nova/+bug/1452568 | 07:45 |
openstack | Launchpad bug 1452568 in OpenStack Compute (nova) "nova allows to live-migrate instance from one availability zone to another" [Undecided,New] | 07:45 |
*** pask81 has joined #openstack-nova | 07:45 | |
*** VW_ has quit IRC | 07:46 | |
*** yingjun has quit IRC | 07:47 | |
*** gszasz has joined #openstack-nova | 07:52 | |
*** mmedvede has joined #openstack-nova | 07:53 | |
*** yingjun has joined #openstack-nova | 07:54 | |
*** jistr has joined #openstack-nova | 07:54 | |
openstackgerrit | sahid proposed openstack/nova: servicegroup: remove get_all method never used as public https://review.openstack.org/168876 | 07:56 |
*** achanda has quit IRC | 07:56 | |
openstackgerrit | Feng Xi Yan proposed openstack/nova: Make VMWare Driver support third party distributed switch. https://review.openstack.org/152006 | 07:59 |
*** jlanoux has joined #openstack-nova | 08:00 | |
*** park_heijlong has joined #openstack-nova | 08:00 | |
*** otter768 has joined #openstack-nova | 08:03 | |
*** jcoufal_ has joined #openstack-nova | 08:03 | |
*** jcoufal has quit IRC | 08:07 | |
*** otter768 has quit IRC | 08:08 | |
*** yingjun has quit IRC | 08:14 | |
*** tdurakov has quit IRC | 08:15 | |
*** yingjun has joined #openstack-nova | 08:18 | |
*** atuvenie has joined #openstack-nova | 08:21 | |
*** alanf-mc_ has joined #openstack-nova | 08:22 | |
*** yassine_ has joined #openstack-nova | 08:22 | |
*** jlanoux has quit IRC | 08:23 | |
*** emagana has joined #openstack-nova | 08:26 | |
*** alanf-mc has quit IRC | 08:26 | |
*** jlanoux has joined #openstack-nova | 08:26 | |
*** alanf-mc_ has quit IRC | 08:27 | |
*** lucasagomes has joined #openstack-nova | 08:28 | |
*** emagana has quit IRC | 08:31 | |
*** ZZelle has quit IRC | 08:31 | |
*** zhangtralon has quit IRC | 08:32 | |
*** zhangtralon has joined #openstack-nova | 08:34 | |
*** moshele has quit IRC | 08:34 | |
*** pixelbeat has joined #openstack-nova | 08:34 | |
*** moshele_ has joined #openstack-nova | 08:34 | |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_create/update https://review.openstack.org/160206 | 08:35 |
openstackgerrit | Alex Xu proposed openstack/nova: Remove db layer hard-code permission checks for quota_class_get_all_by_name https://review.openstack.org/160205 | 08:35 |
openstackgerrit | Alex Xu proposed openstack/nova: Cleanup quota_class unittest with appropriate request context https://review.openstack.org/150351 | 08:35 |
*** moshele_ has quit IRC | 08:35 | |
*** moshele has joined #openstack-nova | 08:35 | |
*** atuvenie has quit IRC | 08:38 | |
*** apuimedo has joined #openstack-nova | 08:43 | |
*** danpb has joined #openstack-nova | 08:49 | |
*** dikonoor has quit IRC | 08:51 | |
*** dikonoor has joined #openstack-nova | 08:52 | |
*** markvoelker has joined #openstack-nova | 08:53 | |
*** tdurakov has joined #openstack-nova | 08:55 | |
tdurakov | bauzas, ping | 08:55 |
bauzas | tdurakov: pong | 08:55 |
tdurakov | :) | 08:55 |
*** jcoufal_ has quit IRC | 08:55 | |
bauzas | tdurakov: so I owned your bug since it needs a blueprint actually | 08:55 |
bauzas | tdurakov: what I'm working on | 08:55 |
tdurakov | what is the plan for 1452568? | 08:55 |
*** jcoufal has joined #openstack-nova | 08:56 | |
bauzas | tdurakov: https://review.openstack.org/#/c/179224/ | 08:56 |
tdurakov | ok, so we doesn't allow such migrations at all? | 08:57 |
bauzas | tdurakov: the idea is to verify against the scheduler if the target host is able to handle the vM | 08:58 |
tdurakov | ok, nice) | 08:58 |
*** markvoelker has quit IRC | 08:58 | |
bauzas | tdurakov: there is one flaw tho, that the hints are not persisted | 08:58 |
bauzas | tdurakov: that's another story, that I'm working on too with the help of alaski | 08:59 |
bauzas | tdurakov: so I'm basically owning all the bugs related to that, and I'll just close of all them once the implementation lands | 09:00 |
tdurakov | sounds good) | 09:00 |
*** ildikov has quit IRC | 09:00 | |
*** yamahata has quit IRC | 09:01 | |
tdurakov | bauzas, I'was thinking about default_schedule_zone... here is my point: if i have 2 hosts, one belongs to Az1 another belongs to Az2, there is no-az hosts at all. Instance always belongs to one of these, no matter want it user or not. when instance is booted, after scheduling, we should manually update info in instances table. Thoughts? | 09:10 |
*** romainh has quit IRC | 09:10 | |
*** romainh1 has joined #openstack-nova | 09:10 | |
openstackgerrit | Sylvain Bauza proposed openstack/nova-specs: Proposes to check destination on migrations https://review.openstack.org/179224 | 09:10 |
openstackgerrit | Lingxian Kong proposed openstack/nova: Correct az property of instance when specifying force host/node https://review.openstack.org/163842 | 09:10 |
tdurakov | no matter if default_schedule_zone specified or not | 09:10 |
bauzas | tdurakov: that's a good question | 09:11 |
bauzas | tdurakov: I don't have a clear PoV now | 09:11 |
bauzas | tdurakov: IMHO, instance AZ shouldn't be stored in the instances table, but rather as a request hint | 09:11 |
*** claudiub_ has joined #openstack-nova | 09:12 | |
bauzas | tdurakov: because AZs are related to hosts, not instances | 09:12 |
bauzas | tdurakov: what we call instance AZ is just the wish of the user, that's it, hence a request spec hint | 09:12 |
bauzas | tdurakov: exactly like a scheduler hint | 09:12 |
bauzas | tdurakov: but just saying that requires a plan for fixing all the crap | 09:12 |
bauzas | tdurakov: so, with the help of johnthetubaguy I'm planning to discuss that on the Summit | 09:13 |
tdurakov | i don't think that storing such scheduler hint in instances table is good idea | 09:13 |
bauzas | tdurakov: we won't store the hints in an instances table | 09:13 |
tdurakov | yep | 09:13 |
*** danpb has quit IRC | 09:13 | |
*** dikonoor has quit IRC | 09:14 | |
tdurakov | but from you words it's only hint) | 09:14 |
*** aix has joined #openstack-nova | 09:14 | |
bauzas | tdurakov: just to explain you, we have lots of work already planned for Lemming | 09:14 |
bauzas | eh, Liberty | 09:14 |
tdurakov | )) | 09:14 |
johnthetubaguy | bauzas: we have quite a lot of time for the scheduler, there should be space for that | 09:14 |
johnthetubaguy | with any luck | 09:14 |
bauzas | johnthetubaguy: yup, the main issue is that I need to draw a plan :/ | 09:14 |
*** claudiub has quit IRC | 09:14 | |
bauzas | johnthetubaguy: and I have no specs atm | 09:14 |
tdurakov | i'd prefer to make fix for now | 09:14 |
tdurakov | to update info for table at boot time | 09:15 |
bauzas | johnthetubaguy: re: the AZ issue I mentioned | 09:15 |
*** dikonoor has joined #openstack-nova | 09:15 | |
johnthetubaguy | bauzas: I am drafting out etherpads and things at the moment, hoping to contact folks to assign session leaders etc, later today | 09:15 |
bauzas | johnthetubaguy: I volunteered for the Wed morning sessions | 09:15 |
bauzas | johnthetubaguy: I leave alaski drive the cells v2 stuff tho | 09:15 |
*** ildikov has joined #openstack-nova | 09:16 | |
*** derekh has joined #openstack-nova | 09:16 | |
bauzas | johnthetubaguy: about AZs, I'll try to draw a plan before the summit in an etherpad at least | 09:16 |
BobBall | johnthetubaguy: Hi! Didn't see you there. Do you know who knows the RAX keystone deployment? | 09:16 |
openstackgerrit | Petrut Lucian proposed openstack/nova: Add Hyper-V storage QoS support https://review.openstack.org/175479 | 09:16 |
bauzas | johnthetubaguy: saying exactly what I mentioned above, ie. instance AZ is just a hint | 09:16 |
bauzas | well, should be rather | 09:17 |
bauzas | tdurakov: will you be able to join Vancouver ? | 09:17 |
*** emagana has joined #openstack-nova | 09:20 | |
*** pask81 has quit IRC | 09:20 | |
tdurakov | nope, i miss that | 09:22 |
*** danpb has joined #openstack-nova | 09:23 | |
bauzas | tdurakov: okay, I'll share your voice then | 09:23 |
bauzas | tdurakov: as we were discussing with johnthetubaguy, there will be some etherpad referring to the sessions | 09:23 |
johnthetubaguy | BobBall: yes, but they are not awake, currently we don't use keystone | 09:24 |
tdurakov | i'll push patch for review to update instance info in db with actual az | 09:24 |
tdurakov | bauzas, could you share it with me? | 09:24 |
bauzas | tdurakov: the ones related to the scheduler should happen on Wed morning PST | 09:24 |
BobBall | johnthetubaguy: Well I'm using it to authenticate against Rackspace... :) | 09:24 |
bauzas | tdurakov: that's what johnthetubaguy said, this is not exactly done yet - we basically have to fill in them | 09:24 |
*** emagana has quit IRC | 09:25 | |
johnthetubaguy | BobBall: yeah, good luck with that, what problem are you seeing / what are you trying? | 09:26 |
tdurakov | ok) share it with me as soon as it will be ready | 09:26 |
johnthetubaguy | tdurakov: bauzas: its here, just not very useful right now: https://wiki.openstack.org/wiki/Summit/Liberty/Etherpads#Nova | 09:26 |
bauzas | tdurakov: eh, tbh it will be incrementally filled till the session :) | 09:27 |
BobBall | johnthetubaguy: See PM | 09:27 |
bauzas | johnthetubaguy: yup, saw that yesterday night | 09:27 |
openstackgerrit | zhangtralon proposed openstack/nova: rename the opt scheduler_default_filters https://review.openstack.org/180910 | 09:28 |
openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: hypervisor support matrix: add feature "serial console" https://review.openstack.org/180912 | 09:31 |
*** yingjun has quit IRC | 09:32 | |
*** park_heijlong has quit IRC | 09:33 | |
tdurakov | bauzas, btw, I've checked how volume attachments work with AZ: if instance located in one az, and volume created in another one. volume attachment won't work | 09:34 |
bauzas | tdurakov: IIRC there is a conf flag for that | 09:34 |
tdurakov | you could even not specify az while creating volume | 09:34 |
tdurakov | i know | 09:34 |
tdurakov | cross_az | 09:34 |
tdurakov | I'm about design | 09:35 |
tdurakov | in cinder volume couldn't belong to no az | 09:35 |
tdurakov | there is also one | 09:36 |
*** dims has joined #openstack-nova | 09:36 | |
tdurakov | and that's ok for me | 09:36 |
tdurakov | bauzas, johnthetubaguy, I'll prefer to implement same for instances | 09:37 |
bauzas | tdurakov: I don't get your point | 09:37 |
bauzas | tdurakov: what do you want ? | 09:37 |
tdurakov | to populate instances db table with correct data. | 09:38 |
tdurakov | wanted to hear you opinion | 09:38 |
johnthetubaguy | tdurakov: I don't have enough context without going digging, but that sounds quite sensible | 09:39 |
tdurakov | give me 5 minutes | 09:40 |
tdurakov | i'll push patch | 09:40 |
*** dims has quit IRC | 09:41 | |
*** redbeard has joined #openstack-nova | 09:41 | |
openstackgerrit | Roman Podoliaka proposed openstack/nova: Fix rebuild of an instance with a volume attached https://review.openstack.org/176891 | 09:42 |
bauzas | tdurakov: MHO is that we should remove the AZ field in the instances table | 09:43 |
bauzas | tdurakov: and persist it in the request_spec | 09:43 |
*** tojuvone has joined #openstack-nova | 09:43 | |
openstackgerrit | John Garbutt proposed openstack/nova: devref: add information to clarify nova scope https://review.openstack.org/178623 | 09:43 |
bauzas | eh, YVR titling all the etherpads, are we meeting in the airport, and so in which gate ? :D | 09:44 |
openstackgerrit | John Garbutt proposed openstack/nova: devref: add information to clarify nova scope https://review.openstack.org/178623 | 09:45 |
bauzas | tdurakov: there are already some changes related to that, but I think we need a better strategy | 09:46 |
openstackgerrit | Timofey Durakov proposed openstack/nova: Fix for availability_zone in instances table https://review.openstack.org/180915 | 09:47 |
*** deepthi has quit IRC | 09:48 | |
*** pask81 has joined #openstack-nova | 09:48 | |
*** IanGovett has joined #openstack-nova | 09:48 | |
openstackgerrit | John Garbutt proposed openstack/nova: devref: update the nova architecture doc https://review.openstack.org/177374 | 09:49 |
bauzas | tdurakov: I just made a few comments in https://etherpad.openstack.org/p/YVR-nova-scheduler-in-liberty | 09:50 |
*** abhishekk has joined #openstack-nova | 09:50 | |
openstackgerrit | John Garbutt proposed openstack/nova: devref: add information to clarify nova scope https://review.openstack.org/178623 | 09:50 |
*** markvoelker has joined #openstack-nova | 09:54 | |
lxsli | bauzas: YVR? | 09:54 |
bauzas | lxsli: airport codename AFAIK | 09:55 |
bauzas | lxsli: maybe just a pirouette for avoiding to mention the liberty gate :) | 09:55 |
openstackgerrit | garyk proposed openstack/nova: VMware: add in vnic resource limitations https://review.openstack.org/180202 | 09:55 |
openstackgerrit | garyk proposed openstack/nova: VMware: create commom object for instance Limits https://review.openstack.org/179059 | 09:55 |
openstackgerrit | garyk proposed openstack/nova: VMware: Resource limits for memory https://review.openstack.org/179060 | 09:55 |
openstackgerrit | garyk proposed openstack/nova: VMware: store extra_specs object https://review.openstack.org/179061 | 09:55 |
openstackgerrit | garyk proposed openstack/nova: VMware: add resource limits for disk https://review.openstack.org/179062 | 09:55 |
bauzas | aka. lemminggate :) | 09:55 |
tdurakov | bauzas, AZs are host related, not instances - that's good. | 09:57 |
bauzas | tdurakov: the only reason I see for persisting the instance AZ information is when migrating | 09:57 |
tdurakov | instance AZ is actually a hint given by the user, so that's a request_spec information - as discussed before, that's confusing | 09:58 |
tdurakov | don't like 'hint' | 09:58 |
*** markvoelker has quit IRC | 09:58 | |
tdurakov | it's strange to store hints there | 09:59 |
bauzas | tdurakov: I don't want to paint my shed, but it is - that's a request data | 09:59 |
bauzas | tdurakov: the word 'hint' is maybe confusing, agreed | 09:59 |
bauzas | tdurakov: technically, that's a SLA | 09:59 |
*** dims_ has joined #openstack-nova | 09:59 | |
tdurakov | :) i'll prefer to use this field to store actual info | 09:59 |
*** lucasagomes_ has joined #openstack-nova | 10:00 | |
bauzas | tdurakov: you mean the instance.az field ? it should tell us bye-bye IMHO | 10:00 |
tdurakov | for now | 10:00 |
tdurakov | until it say good bye) | 10:00 |
*** dulek has quit IRC | 10:01 | |
bauzas | tdurakov: as I mentioned, there are already a few changes working on that, sec | 10:01 |
bauzas | tdurakov: https://review.openstack.org/#/c/163842/ is one of those | 10:01 |
bauzas | tdurakov: just giving you idea on how hacky this is | 10:01 |
bauzas | tdurakov: because we also cache this information | 10:01 |
*** lucasagomes has quit IRC | 10:02 | |
bauzas | https://review.openstack.org/#/c/149562/ and https://review.openstack.org/#/c/172906/ too | 10:02 |
*** dulek has joined #openstack-nova | 10:02 | |
bauzas | tdurakov: I'm sorry, I'll have to stop this convo - really useful because you helped me thinking about the real issue | 10:02 |
*** IanGovett1 has joined #openstack-nova | 10:02 | |
bauzas | tdurakov: now I have an opinion | 10:03 |
bauzas | tdurakov: but I have lots of things to do as well, so I need to stop :) | 10:03 |
*** dulek has quit IRC | 10:03 | |
tdurakov | ok, thanks. it was helpful | 10:03 |
bauzas | as well, | 10:04 |
bauzas | bye | 10:04 |
*** otter768 has joined #openstack-nova | 10:04 | |
*** IanGovett has quit IRC | 10:05 | |
*** zhangtralon has quit IRC | 10:06 | |
*** lucasagomes_ is now known as lucasagomes | 10:08 | |
*** otter768 has quit IRC | 10:09 | |
*** romainh1 has quit IRC | 10:13 | |
*** emagana has joined #openstack-nova | 10:14 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 10:18 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 10:18 |
*** emagana has quit IRC | 10:18 | |
*** liusheng has quit IRC | 10:19 | |
*** moshele has quit IRC | 10:19 | |
*** liusheng has joined #openstack-nova | 10:21 | |
*** moshele has joined #openstack-nova | 10:21 | |
*** ildikov has quit IRC | 10:22 | |
openstackgerrit | Timofey Durakov proposed openstack/nova: Fix for availability_zone in instances table https://review.openstack.org/180915 | 10:37 |
*** mpsairam has quit IRC | 10:38 | |
*** vladikr has quit IRC | 10:38 | |
*** ildikov has joined #openstack-nova | 10:38 | |
*** coolsvap is now known as coolsvap|afk | 10:43 | |
*** baoli has joined #openstack-nova | 10:43 | |
*** ociuhandu has joined #openstack-nova | 10:47 | |
*** baoli has quit IRC | 10:48 | |
*** alexpilotti has joined #openstack-nova | 10:49 | |
bauzas | johnthetubaguy: quick q, is it you who removed my name from the lib summit ideas etherpad or is it my mobile device who said to me I was adding some lines to the etherpad but didn't ? | 10:51 |
bauzas | johnthetubaguy: the etherpad timeslider is buggy for me, can't see what happened and why | 10:51 |
bauzas | johnthetubaguy: (speaking about the volunteers for the sessions) | 10:52 |
johnthetubaguy | bauzas: your name is there as a joint leader I think | 10:54 |
bauzas | johnthetubaguy: strange ? | 10:54 |
bauzas | johnthetubaguy: can't see it | 10:54 |
johnthetubaguy | bauzas: line 15? | 10:54 |
bauzas | johnthetubaguy: I honestly don't take much of importance to be seen as leader, I just want to understand why I thought I wrote something than I can't see it | 10:55 |
*** markvoelker has joined #openstack-nova | 10:55 | |
bauzas | johnthetubaguy: speaking of https://etherpad.openstack.org/p/liberty-nova-summit-ideas | 10:55 |
bauzas | johnthetubaguy: L331 I mentioned my name | 10:55 |
bauzas | johnthetubaguy: but can't see it | 10:56 |
johnthetubaguy | bauzas: oh right, not sure I saw your name there before, my bad | 10:56 |
bauzas | johnthetubaguy: again, that's not much about myself, that's more about if I can trust my android | 10:56 |
bauzas | johnthetubaguy: ok then my android chrome sucks | 10:56 |
johnthetubaguy | bauzas: it could have been me | 10:57 |
johnthetubaguy | bauzas: not yet sure about that separate resource tracker slot, need to see what else crops up, I think there was a suggestion on the ML I need to dig into | 10:57 |
bauzas | johnthetubaguy: yup, I'm thinking about that too | 10:57 |
*** sahid has quit IRC | 10:57 | |
johnthetubaguy | bauzas: anyways, I added into the first scheduler session for now | 10:58 |
bauzas | johnthetubaguy: I would see it as a free slot for discussing urgent queries | 10:58 |
bauzas | johnthetubaguy: cool, that's fine by me | 10:58 |
johnthetubaguy | bauzas: we already have two of those | 10:58 |
*** sahid has joined #openstack-nova | 10:58 | |
*** Longgeek has quit IRC | 10:58 | |
bauzas | johnthetubaguy: yeah, I don't get much importance of the 1150 slot for the scheduler, IMHO we have enough time | 10:58 |
*** Longgeek has joined #openstack-nova | 10:58 | |
bauzas | s/get/give | 10:58 |
johnthetubaguy | bauzas: might just end up being cells/scheduler overflow, but still actively looking for anything else | 10:58 |
bauzas | johnthetubaguy: yup, consider it as free | 10:59 |
bauzas | johnthetubaguy: alaski was maybe interested in hijacking that slot | 10:59 |
johnthetubaguy | bauzas: it could include persistent resource tracker, committing resources during the scheduling process, and general simplification plans and hows doing it | 10:59 |
*** markvoelker has quit IRC | 11:00 | |
bauzas | johnthetubaguy: scaling out the scheduler for cells v2 is IMHO a huge discussion | 11:00 |
bauzas | johnthetubaguy: so it could probably just leave the other cells v2 topic undiscussed | 11:00 |
bauzas | topic* | 11:00 |
bauzas | topics even | 11:00 |
bauzas | but we have a 2nd slot for cells V2 | 11:00 |
bauzas | so, anyway, that's your call, take it as your convenience | 11:01 |
bauzas | I'm fine with sticking with what we have already, the 1150 slot is just bonus to me | 11:01 |
bauzas | either for cells or scheduler | 11:01 |
openstackgerrit | garyk proposed openstack/nova: VMware: support limits via image meta data https://review.openstack.org/180942 | 11:01 |
bauzas | johnthetubaguy: ^ | 11:01 |
*** achanda has joined #openstack-nova | 11:02 | |
johnthetubaguy | bauzas: yeah, lets see what comes up | 11:02 |
bauzas | johnthetubaguy: sure, PCI and NUMA sounds also a good fit btw. | 11:03 |
bauzas | johnthetubaguy: just because we made actions during the last Summit about 3rd party CI etc. | 11:03 |
bauzas | johnthetubaguy: and I would love having feedback on that and if any other actionable items to do | 11:04 |
johnthetubaguy | bauzas: yes thats true, I added PCI hardening to that slot | 11:04 |
bauzas | johnthetubaguy: IIRC there is a proposal for a NUMA priority for Lib | 11:05 |
johnthetubaguy | bauzas: actually those are the key issues with the resource tracker, so thats maybe a good way to twist that | 11:05 |
bauzas | johnthetubaguy: lemme check | 11:05 |
bauzas | johnthetubaguy: yup exactly, we suffer of that | 11:05 |
bauzas | johnthetubaguy: yeah, NFV is proposed as a prio, I don't want to talk about that but rather the points it raises | 11:06 |
bauzas | johnthetubaguy: https://etherpad.openstack.org/p/liberty-nova-priorities mentions CI issues, distro problems etc. | 11:07 |
bauzas | johnthetubaguy: sounds enough chatty for one session, if we cover PCI and NUMA IMHO | 11:07 |
johnthetubaguy | yeah | 11:07 |
*** achanda has quit IRC | 11:07 | |
johnthetubaguy | I like that | 11:07 |
johnthetubaguy | well, I don't, but its a good idea, if we get the correct discussion going | 11:07 |
johnthetubaguy | bauzas: can you check these two out for me now please: https://etherpad.openstack.org/p/YVR-nova-scalling-out-scheduler-for-cells https://etherpad.openstack.org/p/YVR-nova-scheduler-in-liberty | 11:08 |
bauzas | johnthetubaguy: yeah, it's all about seeing what's worth doing | 11:08 |
bauzas | johnthetubaguy: sure will do | 11:08 |
bauzas | johnthetubaguy: I already noted a few things for the latter | 11:08 |
*** emagana has joined #openstack-nova | 11:08 | |
bauzas | johnthetubaguy: and I'll look at the former | 11:09 |
*** ankit_ag has quit IRC | 11:11 | |
*** ankit_ag has joined #openstack-nova | 11:11 | |
*** emagana has quit IRC | 11:13 | |
*** bfic has joined #openstack-nova | 11:15 | |
*** VW_ has joined #openstack-nova | 11:18 | |
*** VW_ has quit IRC | 11:22 | |
*** kaufer has joined #openstack-nova | 11:25 | |
openstackgerrit | Lianhao Lu proposed openstack/nova-specs: Add new API to list compute node metrics https://review.openstack.org/180049 | 11:26 |
*** oro has joined #openstack-nova | 11:26 | |
*** ferest has joined #openstack-nova | 11:30 | |
*** links has quit IRC | 11:31 | |
*** moshele has quit IRC | 11:32 | |
*** ferest has quit IRC | 11:32 | |
*** links has joined #openstack-nova | 11:32 | |
*** sudipto has joined #openstack-nova | 11:34 | |
*** dulek has joined #openstack-nova | 11:38 | |
*** moshele has joined #openstack-nova | 11:41 | |
*** moshele has quit IRC | 11:42 | |
*** romainh has joined #openstack-nova | 11:42 | |
openstackgerrit | Sean Dague proposed openstack/nova: don't report service group connection events as errors https://review.openstack.org/180956 | 11:45 |
sdague | so... I'm sifting through some of these fails during upgrades, and nova-compute is barf o matic during the partial upgrade | 11:50 |
*** sudipto has quit IRC | 11:50 | |
sdague | because the periodic tasks to do things with the db try to run when conductor is down | 11:50 |
sdague | have we considered that window, or are we just hoping that compute won't do anything bad with partial data | 11:51 |
*** openstackgerrit has quit IRC | 11:52 | |
*** openstackgerrit has joined #openstack-nova | 11:53 | |
*** VW_ has joined #openstack-nova | 11:53 | |
*** markvoelker has joined #openstack-nova | 11:56 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 11:56 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 11:56 |
*** VW_ has quit IRC | 11:57 | |
*** mmedvede_ has joined #openstack-nova | 11:59 | |
*** kiran-r has quit IRC | 12:00 | |
*** markvoelker has quit IRC | 12:00 | |
dikonoor | alex_xu : Hi | 12:01 |
dikonoor | alex_xu: Do you know when will nova start moving to use oslo_policy? | 12:02 |
*** emagana has joined #openstack-nova | 12:02 | |
dikonoor | alex_xu : I have some very basic questions around the policy enforcement and the more i look at it , the more confused I am . I am trying to take a closer look at what really needs to be passed to policy.enforce to ensure that the ownership rules like "owner": "user_id:%(user_id)s" work | 12:04 |
*** raildo has joined #openstack-nova | 12:05 | |
*** otter768 has joined #openstack-nova | 12:05 | |
*** yassine_ has quit IRC | 12:05 | |
dikonoor | alex_xu: If I am not wrong , the enforce method expects the rule name , context/credentials from the request you are trying to make and a dictionary around the target object..The dictionary around the target object is what confuses me | 12:05 |
*** emagana has quit IRC | 12:07 | |
dikonoor | alex_xu : I was initially under the impression that the enforce method basically tries to match the contents from the request credentials to the target and thus attempts to figure out if the user making the request has the right ownership on the target resources (based on the rule defined in the policy.json) | 12:07 |
*** raildo has quit IRC | 12:10 | |
*** otter768 has quit IRC | 12:10 | |
*** raildo has joined #openstack-nova | 12:10 | |
dikonoor | alex_xu: and probably that's what it does.. but I see that in most of the nova api calls the authorize call (which in turn calls policy.enforce) is called right at the beginning of the api method..So , if we take a case where I want only the resource owner (i e user) to update the resource , the call to authorize with not carry any value for target.. | 12:10 |
*** doude_ has quit IRC | 12:10 | |
*** doude has joined #openstack-nova | 12:11 | |
*** ankit_ag has quit IRC | 12:11 | |
*** ankit_ag has joined #openstack-nova | 12:12 | |
dikonoor | alex_xu : As the target is None , target is set to some values pulled from the context >> https://github.com/openstack/nova/blob/master/nova/api/openstack/extensions.py#L341 | 12:12 |
dikonoor | alex_xu: so where does the filtering that this resource is owned by not by the user making the request happen? | 12:13 |
*** markvoelker has joined #openstack-nova | 12:15 | |
dikonoor | dhellmann : alex_xu: could you guys (or someone else) point me to the code where that logic runs ..i e in the flow I mentioned here , where does it try to figure out if the resource is owned by the user making the request | 12:16 |
*** dprince has joined #openstack-nova | 12:16 | |
*** liusheng has quit IRC | 12:16 | |
*** aysyd has joined #openstack-nova | 12:18 | |
*** liusheng has joined #openstack-nova | 12:19 | |
*** dave-mccowan has joined #openstack-nova | 12:24 | |
*** busterswt has joined #openstack-nova | 12:26 | |
*** rfolco has joined #openstack-nova | 12:29 | |
openstackgerrit | jichenjc proposed openstack/nova-specs: Allow adjust ephemeral disks during resize https://review.openstack.org/145736 | 12:31 |
openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: add in vnic resource limitations https://review.openstack.org/180202 | 12:33 |
*** sudipto has joined #openstack-nova | 12:37 | |
*** VW_ has joined #openstack-nova | 12:38 | |
*** bkopilov has quit IRC | 12:39 | |
alex_xu | dikonoor: currently we didn't enable user owned permission, I plan to do that in the beginning, but unfortunately that didn't include in the initial policy spec. And if we want to do that, it need some behavior change for the API. | 12:40 |
alex_xu | dikonoor: target with the resource obj only bring to us one benefit is enable user-owned permission, I'm not sure is there anyone using nova like that, so I list that works as low priority | 12:41 |
*** bfic has quit IRC | 12:44 | |
dikonoor | alex_xu: That's very interesting because I tried few with ownerhsip rules and it seems to work and I thought it was already implemented.. | 12:45 |
dikonoor | alex_xu ..For eg. I tried with these >> "compute:start": "rule:admin_or_owner", | 12:45 |
dikonoor | "compute:stop": "rule:admin_or_owner", | 12:45 |
dikonoor | "compute:pause": "rule:admin_or_owner", | 12:45 |
dikonoor | "compute:unpause": "rule:admin_or_owner", | 12:45 |
dikonoor | "compute:update": "rule:admin_or_owner", | 12:45 |
*** bfic has joined #openstack-nova | 12:45 | |
alex_xu | dikonoor: ok, I see what you mean now... | 12:45 |
dikonoor | alex_xu : with rule defined as >> "owner": "user_id:%(user_id)s", | 12:45 |
dikonoor | "admin_or_owner": "rule:adm or rule:owner", | 12:45 |
*** bfic has quit IRC | 12:45 | |
alex_xu | dikonoor: that works even without target | 12:45 |
alex_xu | dikonoor: the owner permission is done by db code | 12:46 |
dikonoor | alex_xu : Ah ..ok..db layer.. | 12:46 |
dikonoor | alex_xu : So the db code does that layer of filter based on the ownership rule defined in policy.json? Could you point me to some code that does that db filtering? | 12:47 |
*** yassine_ has joined #openstack-nova | 12:49 | |
alex_xu | dikonoor: ok, just saw you said user_id:%(user_id)s, the target is passed at here https://github.com/openstack/nova/blob/master/nova/compute/api.py#L225 | 12:49 |
openstackgerrit | garyk proposed openstack/nova: VMware: support limits via image meta data https://review.openstack.org/180942 | 12:49 |
dikonoor | alex_xu : Apart from the lifecycle operations , I have tried this in others like attach_volume , detach_volume and many others..The rules "owner": "user_id:%(user_id)s" or "owner": "project_id:%(project_id)s" works for all these | 12:49 |
*** dulek has quit IRC | 12:50 | |
*** dims_ has quit IRC | 12:51 | |
*** dims has joined #openstack-nova | 12:51 | |
*** jichen has quit IRC | 12:52 | |
*** dulek has joined #openstack-nova | 12:54 | |
dikonoor | alex_xu : these have targets..I agree..When you said earlier that the ownerhsip stuff is not implemented yet , I guess you meant something else..what was it? | 12:55 |
alex_xu | dikonoor: 'compute:*' is compute api layer, the api layer without the target | 12:56 |
alex_xu | dikonoor: "os_compute_api:*" won't support user-owned permission | 12:56 |
alex_xu | dikonoor: and "compute_extension:*" won't support also | 12:57 |
*** emagana has joined #openstack-nova | 12:57 | |
*** signed8bit_ZZZzz has quit IRC | 12:57 | |
*** IanGovett1 has quit IRC | 12:58 | |
alex_xu | dikonoor: do you have real use-case for user want to use user-owned permission? | 12:59 |
RuiChen | johnthetubaguy: can you have a look https://review.openstack.org/#/c/169638/ when you have time? please feel free for some comments, thanks. | 12:59 |
*** claudiub_ has quit IRC | 13:00 | |
*** IanGovett has joined #openstack-nova | 13:00 | |
*** ildikov has quit IRC | 13:00 | |
*** moshele has joined #openstack-nova | 13:00 | |
*** emagana has quit IRC | 13:01 | |
openstackgerrit | sahid proposed openstack/nova-specs: console: convert consoles code to use objects framework https://review.openstack.org/165838 | 13:01 |
*** mriedem has joined #openstack-nova | 13:02 | |
*** ildikov has joined #openstack-nova | 13:02 | |
*** Mike_D_laptop has joined #openstack-nova | 13:04 | |
dikonoor | dikonoor: I wasn't aware that there were target based differences across diferent api layers within nova..lemme go through it a little more. yeah.. user and project-based ownerhsip is something that all serious cloud users ask for /would want to have.. | 13:05 |
sahid | dansmith: any chance to get your review here https://review.openstack.org/#/c/171287/ ? | 13:06 |
openstackgerrit | Sudipta Biswas proposed openstack/nova-specs: Report host memory b/w as a metric in Nova https://review.openstack.org/180983 | 13:06 |
*** yingjun has joined #openstack-nova | 13:07 | |
*** neelashah has joined #openstack-nova | 13:08 | |
mriedem | sahid: is there some other change that's part of https://blueprints.launchpad.net/nova/+spec/resource-objects that will use that field? | 13:08 |
*** tonytan4ever has quit IRC | 13:08 | |
sahid | mriedem: looking | 13:09 |
mriedem | sahid: like one of these maybe https://review.openstack.org/#/q/topic:resource-objects,n,z | 13:09 |
mriedem | sahid: i'd ask jaypipes on that before removing it | 13:09 |
*** IanGovett has quit IRC | 13:10 | |
sahid | mriedem: yes it probably better to get te return from jaypipes I looked at when it has been introduced | 13:10 |
sahid | but sure more reviews we have better it will be | 13:11 |
*** baoli_ has joined #openstack-nova | 13:11 | |
*** kiran-r has joined #openstack-nova | 13:12 | |
alex_xu | dikonoor: ok, let me think about it. | 13:13 |
*** baoli_ has quit IRC | 13:13 | |
*** IanGovett has joined #openstack-nova | 13:13 | |
*** burt has joined #openstack-nova | 13:13 | |
*** baoli has joined #openstack-nova | 13:14 | |
alex_xu | advertisement: the policy patches related to service, compute_node and quota are in good shape. Appreciate any review! https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/nova-api-policy-final-part,n,z | 13:15 |
openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: Hypervisor Support Matrix renders links in notes https://review.openstack.org/180985 | 13:16 |
*** tjones1 has joined #openstack-nova | 13:16 | |
*** tjones1 has left #openstack-nova | 13:17 | |
*** david-lyle has quit IRC | 13:19 | |
*** park_heijlong has joined #openstack-nova | 13:20 | |
openstackgerrit | sahid proposed openstack/nova: compute: make test rpcapi more flexible https://review.openstack.org/164177 | 13:21 |
*** yingjun has quit IRC | 13:23 | |
*** cfriesen has joined #openstack-nova | 13:23 | |
*** dboik has quit IRC | 13:24 | |
*** openstackstatus has quit IRC | 13:25 | |
sahid | mriedem: btw :) if i can ask you.. we have two backports with clean cherry-pick can you take a look at them? | 13:25 |
sahid | https://review.openstack.org/#/c/180075/ | 13:25 |
sahid | https://review.openstack.org/#/c/178072/ | 13:25 |
mriedem | i can | 13:26 |
sahid | garyk: already acked the first one and I guess he will be happy to ack the other then | 13:26 |
*** openstackstatus has joined #openstack-nova | 13:26 | |
*** ChanServ sets mode: +v openstackstatus | 13:26 | |
sahid | great thank you | 13:26 |
*** eharney has quit IRC | 13:28 | |
openstackgerrit | Anne Gentle proposed openstack/nova: Changes conf.py for Sphinx build because oslosphinx now contains GA https://review.openstack.org/172525 | 13:29 |
garyk | mriedem: without it a large majority of instances do not boot | 13:30 |
mriedem | garyk: so where is the bug? | 13:30 |
mriedem | a backport w/o a bug reference is suspect | 13:30 |
dansmith | sahid: just FYI, in general this sort of thing is not very useful right now since we've moved everything to o.vo and are going to drop all of nova's stuff like this anyway | 13:31 |
dansmith | sahid: but, this one isn't in o.vo, so I guess it's okay to remove | 13:31 |
*** kaufer has quit IRC | 13:32 | |
*** changbl has quit IRC | 13:33 | |
*** gokrokve has joined #openstack-nova | 13:33 | |
garyk | i was sure that i added a bug. | 13:33 |
sahid | dansmith: yes i understand, thanks for the information | 13:34 |
mriedem | alexpilotti: are you around to talk about this hyper-v gen2 thing in the mailing list? | 13:35 |
*** moshele has quit IRC | 13:37 | |
*** yamahata has joined #openstack-nova | 13:37 | |
*** moshele has joined #openstack-nova | 13:37 | |
openstackgerrit | ChangBo Guo(gcb) proposed openstack/nova: Filter out inactive compute firstly https://review.openstack.org/180155 | 13:40 |
*** park_heijlong has quit IRC | 13:42 | |
*** signed8bit has joined #openstack-nova | 13:42 | |
openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: handle NotSupportedError in compareCPU https://review.openstack.org/166130 | 13:42 |
*** vladikr has joined #openstack-nova | 13:43 | |
sdague | dansmith: so... in looking into some of the more recent grenade failures on the partial job, once we've dropped conductor, the compute periodic jobs on the old side keep chugging | 13:44 |
sdague | and mostly stack tracing | 13:44 |
sdague | because the idea that conductor might be down is... never anticipated | 13:44 |
dansmith | sdague: so my change hit that same thing again.. have you seen it on other jobs, or is that somethign else? | 13:44 |
*** dboik has joined #openstack-nova | 13:44 | |
dansmith | sdague: right, they've always done that, AFAIK | 13:44 |
*** dboik has quit IRC | 13:44 | |
sdague | so I've seen that on a bunch of jobs | 13:45 |
dansmith | which, the stack tracing or the thing I was hitting? | 13:45 |
*** dboik has joined #openstack-nova | 13:45 | |
sdague | I'm seeing the stack tracing a lot | 13:45 |
*** signed8b_ has joined #openstack-nova | 13:45 | |
*** zzzeek has joined #openstack-nova | 13:46 | |
sdague | I'm seeing the computes loose networking some times | 13:46 |
dansmith | right, that should have been happening since the beginning of the partial job | 13:46 |
dansmith | the stack-tracing that is | 13:46 |
sdague | ok, that seems, suboptimal right, because are we sure the periodic tasks are atomic? (i.e. won't mess anything up if they lose connectivity part way through) | 13:47 |
dansmith | well, it's the same as any other sort of connectivity loss, right? | 13:47 |
mriedem | i don't think we are sure of that | 13:47 |
*** kaufer1 has joined #openstack-nova | 13:47 | |
dansmith | and/or just a regular timeout to do a thing | 13:47 |
mriedem | sync_instance_power_state could be sending a stop API that fails in the middle b/c conductor is down | 13:48 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 13:48 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 13:48 |
dansmith | if you request a thing and then another thing, those go to two different conductors, | 13:48 |
dansmith | and if one is wedged, then you'll timeout in the same way | 13:48 |
dansmith | so that problem isn't really any different than normal runtime | 13:48 |
openstackgerrit | dave-mcnally proposed openstack/nova: Add policy to cover snapshotting of volume backed instances https://review.openstack.org/176805 | 13:48 |
danpb | dansmith: i'm looking at the Boolean field type and its coerce() method which simply does 'bool(value)' | 13:49 |
dansmith | mriedem: what does the stop example matter? | 13:49 |
*** gokrokve has quit IRC | 13:49 | |
danpb | dansmith: this means that strings like 'false', 'no' all end up being coerced to True | 13:49 |
*** signed8bit has quit IRC | 13:49 | |
*** r-daneel has joined #openstack-nova | 13:49 | |
dansmith | danpb: yeah, it's truthiness, and any non-zero string is truthy | 13:49 |
danpb | dansmith: right but for the image properties object we really need it to do strutils.bool_from_string() | 13:50 |
mriedem | dansmith: it doesn't really, just means the periodic tasks shouldn't expect conductor/api to be up | 13:50 |
danpb | so i'm wondering whether changing the Boolean field to use bool_from_string is acceptable, | 13:50 |
danpb | or whether we need todo somethin different | 13:50 |
sdague | dansmith: well I'm trying to poke at reasons why a guest on the network was responding to pings, 0 api calls were made, then it stops responding to pings | 13:50 |
dansmith | mriedem: well everything they do with checking/fixing state requires conductor to be up, but I agree they should be idempotent | 13:50 |
*** kiran-r has quit IRC | 13:51 | |
dansmith | sdague: yeah.. do any of the periodics affect firewall rules? I can't think of any off the top of my head | 13:51 |
*** jamielennox is now known as jamielennox|away | 13:51 | |
mriedem | dansmith: umm | 13:51 |
dansmith | danpb: it's part of the API of the library now, so you can't really make that change, no | 13:51 |
dansmith | danpb: you can add a LooseBool field that inherits from bool that coerces with that though and use that in your object | 13:52 |
mriedem | there is the heal_instance_info_cache task, that doesn't directly do anything with fw rules, but wondering if something bad could happen there | 13:52 |
mriedem | loosebool sounds too much like loosestoole | 13:52 |
mriedem | *stool | 13:52 |
dansmith | mriedem: I don't think so, that one is mostly just calling to neutron, storing the result, IIRC | 13:52 |
dansmith | mriedem: danpb can feel free to come up with a less scatalogical name if he prefers :) | 13:53 |
*** gokrokve has joined #openstack-nova | 13:53 | |
danpb | dansmith: ok, i'll do that approach | 13:53 |
*** gszasz has quit IRC | 13:54 | |
*** links has quit IRC | 13:54 | |
*** gokrokve has quit IRC | 13:54 | |
*** IanGovett1 has joined #openstack-nova | 13:55 | |
*** gokrokve has joined #openstack-nova | 13:55 | |
*** oro has quit IRC | 13:55 | |
dansmith | cool | 13:56 |
*** gszasz has joined #openstack-nova | 13:57 | |
*** IanGovett has quit IRC | 13:57 | |
*** zul has quit IRC | 13:57 | |
*** sudipto has quit IRC | 13:59 | |
sdague | ok, so we've got 16 fails in the last 7 days on the test servers not pinging - http://logstash.openstack.org/#eyJzZWFyY2giOiJcImNvdWxkbid0IHBpbmdcIiIsImZpZWxkcyI6W10sIm9mZnNldCI6MCwidGltZWZyYW1lIjoiNjA0ODAwIiwiZ3JhcGhtb2RlIjoiY291bnQiLCJ0aW1lIjp7InVzZXJfaW50ZXJ2YWwiOjB9LCJzdGFtcCI6MTQzMTAwNjg2OTUzOH0= | 14:01 |
sahid | jogo: funny here https://review.openstack.org/#/c/164168/ ;) | 14:01 |
sdague | they don't all look the same, but I'm trying to figure out what would be the best additional information to get to the bottom of these | 14:02 |
*** zul has joined #openstack-nova | 14:02 | |
sdague | I've got code going in that will give us iptables dumps on fails, but perhaps also doing those dumps at various stages could help figure out working vs. non. | 14:03 |
*** thangp has joined #openstack-nova | 14:03 | |
*** eglynn has joined #openstack-nova | 14:03 | |
mriedem | sdague: these all look like nova-network jobs? | 14:04 |
dansmith | sdague: I've got a call now, but can jump back in in an hour | 14:04 |
sdague | mriedem: yep | 14:04 |
mriedem | sdague: hmm | 14:04 |
*** nkrinner has quit IRC | 14:04 | |
mriedem | sdague: would dumping the network info_cache be any help? | 14:04 |
garyk | johnthetubaguy: can you please post the link for the etherpad for the sessions | 14:04 |
sdague | don't know, I'm happy to take options here :) | 14:04 |
mriedem | we log the nw info_cache in the n-net logs everywhere but it's a big ugly dict so it's a bit hard to trace | 14:05 |
sdague | though it is interesting that neutron is more reliable here | 14:05 |
mriedem | would be nice to know at the point of failure what our nw info_cache looks like compared to iptables | 14:05 |
dansmith | sdague: do we do a partial run with neutron? | 14:05 |
mriedem | yes | 14:05 |
sdague | dansmith: we do not | 14:05 |
sdague | however this isn't only partial fails | 14:05 |
mriedem | isn't there one on experimental? | 14:05 |
dansmith | and this only happens on the partial job right? | 14:05 |
dansmith | ah, okay | 14:05 |
*** achanda has joined #openstack-nova | 14:06 | |
*** otter768 has joined #openstack-nova | 14:06 | |
mriedem | i thought there was a grenade partial neutron on the exp queue | 14:06 |
sdague | so... that signature is just a "stuff doesn't work" signature | 14:06 |
sdague | I think this is at least a couple of different problems overlapping | 14:06 |
mriedem | i'd have to look but, | 14:06 |
*** eharney has joined #openstack-nova | 14:06 | |
mriedem | alex_xu: and jichenjc and i were looking at some n-net errors in some other race bugs | 14:06 |
mriedem | on fixed ip alloc/dealloc | 14:07 |
sdague | mriedem: ok | 14:07 |
mriedem | https://review.openstack.org/#/c/171390/ | 14:07 |
mriedem | https://bugs.launchpad.net/nova/+bug/1249065/comments/30 | 14:07 |
openstack | Launchpad bug 1249065 in OpenStack Compute (nova) "Nova throws 400 when attempting to add floating ip (No nw_info cache associated with instance)" [Medium,In progress] - Assigned to Matt Riedemann (mriedem) | 14:07 |
sdague | yeh, so in all these runs I'm allocating floating ips to the guests because that seemed like the more generic way to handle it regardless of neutron / nova | 14:08 |
alex_xu | mriedem: I didn't take a look at neutron case yet, will try to take a look at it weekend | 14:08 |
openstackgerrit | Rajesh Tailor proposed openstack/nova: Fix quota-update when nova-compute startup finish https://review.openstack.org/122347 | 14:08 |
sdague | because the ovs networks don't let you connect from host to guest | 14:08 |
sdague | mriedem: https://review.openstack.org/#/c/171390 is merge conflicting | 14:08 |
mriedem | i know | 14:09 |
mriedem | lots of my changes are merge conflicting this week | 14:09 |
sdague | heh | 14:09 |
sdague | ok, stepping away for a bit | 14:09 |
alex_xu | johnthetubaguy: jaypipes do you have want an api meeting this week? As Ken'ichi can't ensure join the meeting and gilliard_afk is on vacation, if you didn't have anything want to talk about, then I think we can cancel api meeting this week | 14:10 |
*** oro has joined #openstack-nova | 14:10 | |
*** achanda has quit IRC | 14:11 | |
*** otter768 has quit IRC | 14:11 | |
openstackgerrit | Chris Friesen proposed openstack/nova: unify libvirt driver checks for qemu https://review.openstack.org/175542 | 14:11 |
*** signed8b_ has quit IRC | 14:12 | |
*** jcoufal has quit IRC | 14:14 | |
*** ankit_ag has quit IRC | 14:14 | |
*** abhishekk has quit IRC | 14:15 | |
*** bkopilov has joined #openstack-nova | 14:20 | |
*** baoli has quit IRC | 14:22 | |
*** ddaskal has joined #openstack-nova | 14:22 | |
*** baoli_ has joined #openstack-nova | 14:23 | |
openstackgerrit | Eric Brown proposed openstack/nova: VMware: Allow other nested hypervisors (HyperV) https://review.openstack.org/181018 | 14:24 |
openstackgerrit | Eric Harney proposed openstack/nova: Add error message to failed block device transform https://review.openstack.org/180584 | 14:25 |
*** bfic has joined #openstack-nova | 14:25 | |
*** moshele has quit IRC | 14:25 | |
*** tpeoples has joined #openstack-nova | 14:26 | |
*** signed8bit has joined #openstack-nova | 14:26 | |
*** signed8bit has quit IRC | 14:27 | |
*** jcoufal has joined #openstack-nova | 14:27 | |
tpeoples | Is there a way to dump information about a hung nova-compute process? Telnetting into the eventlet backdoor port just hangs as well. I'd like to be able to determine where the native thread is hung if possible | 14:27 |
mriedem | tpeoples: yeah sec | 14:28 |
mriedem | tpeoples: http://docs.openstack.org/developer/nova/devref/gmr.html | 14:28 |
*** bfic has left #openstack-nova | 14:28 | |
*** achanda has joined #openstack-nova | 14:29 | |
*** achanda has quit IRC | 14:31 | |
jaypipes | alex_xu: I would like to find some resolution on the 403 Forbidden for OverQuota issue with you, sdague, cdent, and others... | 14:31 |
jaypipes | alex_xu: or at least, have a conversation about it, even if no resolution. | 14:32 |
alex_xu | jaypipes: it should on nova-api meeting, or api-wg meeting? | 14:32 |
jaypipes | sahid: https://review.openstack.org/#/c/171287/ approved. thx man. | 14:32 |
tpeoples | mriedem: thanks | 14:33 |
jaypipes | alex_xu: both, probably, but I think either meeting today is bad time for you :( | 14:33 |
openstackgerrit | Eric Brown proposed openstack/nova: VMware: update supported vsphere 6.0 os types https://review.openstack.org/181020 | 14:34 |
alex_xu | jaypipes: yea, api-wg meeting is bad time, but I can try. nova-api meeting is good time for me. | 14:35 |
*** kaufer1 has quit IRC | 14:35 | |
*** Murali has quit IRC | 14:35 | |
jaypipes | alex_xu: ok, cool. let's discuss it then. | 14:36 |
tpeoples | mriedem: does that require me to import gmr and register the hook prior to starting the nova process? I'm not sure if I'll be able to reproduce if I do that (scale testing) so trying to get info about the currently hung process if at all possible | 14:36 |
*** zsais has joined #openstack-nova | 14:37 | |
*** moshele has joined #openstack-nova | 14:38 | |
alex_xu | jaypipes: yea, cool, let me send a note for nova api meeting | 14:39 |
*** zsais22 has joined #openstack-nova | 14:40 | |
*** emagana has joined #openstack-nova | 14:41 | |
openstackgerrit | Eric Brown proposed openstack/nova: VMware: Allow other nested hypervisors (HyperV) https://review.openstack.org/181018 | 14:41 |
sahid | jaypipes: thank you | 14:42 |
*** packet has joined #openstack-nova | 14:43 | |
*** jichen has joined #openstack-nova | 14:43 | |
*** zsais has quit IRC | 14:43 | |
*** tonytan4ever has joined #openstack-nova | 14:45 | |
*** tonytan4ever has quit IRC | 14:49 | |
*** hemnafk is now known as hemna | 14:52 | |
*** zsais222 has joined #openstack-nova | 14:55 | |
*** ociuhandu_ has joined #openstack-nova | 14:56 | |
*** popw has quit IRC | 14:56 | |
*** achanda has joined #openstack-nova | 14:56 | |
*** ociuhandu has quit IRC | 14:56 | |
*** ociuhandu_ is now known as ociuhandu | 14:56 | |
*** xyang1 has joined #openstack-nova | 14:58 | |
*** yamahata has quit IRC | 14:59 | |
*** zsais22 has quit IRC | 14:59 | |
*** baoli_ has quit IRC | 14:59 | |
*** yamahata has joined #openstack-nova | 14:59 | |
*** andrewbogott_afk is now known as andrewbogott | 14:59 | |
*** baoli has joined #openstack-nova | 15:00 | |
dansmith | sdague: mriedem okay, I'm done if I can help further on that network thing | 15:00 |
*** dboik_ has joined #openstack-nova | 15:01 | |
mriedem | dansmith: if you have any ideas about this it might help https://bugs.launchpad.net/nova/+bug/1249065/comments/30 | 15:01 |
openstack | Launchpad bug 1249065 in OpenStack Compute (nova) "Nova throws 400 when attempting to add floating ip (No nw_info cache associated with instance)" [Medium,In progress] - Assigned to Matt Riedemann (mriedem) | 15:01 |
dansmith | mriedem: is that related to the thing sdague was chasing, or something else? | 15:02 |
mriedem | dansmith: well, maybe related | 15:02 |
*** jcoufal has quit IRC | 15:02 | |
mriedem | general n-net funkiness wrt races | 15:02 |
*** achanda has quit IRC | 15:03 | |
mriedem | i'm trying to get my microversion change rebased and fixing the unit tests but the unit tests are throwing me off | 15:03 |
*** mspreitz has joined #openstack-nova | 15:04 | |
*** dboik has quit IRC | 15:05 | |
mriedem | tpeoples: sorry i'm not a gmr wizard, just knew about the link | 15:05 |
dansmith | tpeoples: you send SIGUSR1 to nova-compute and it dumps that data | 15:05 |
*** gszasz has quit IRC | 15:05 | |
dansmith | mriedem: that bug looks pretty bad | 15:05 |
dansmith | mriedem: does it happen right after boot, such that it looks like a race? | 15:06 |
mriedem | dansmith: the floating_ips one? been around for a long time, very rare in the gate | 15:06 |
dansmith | ugh | 15:06 |
tpeoples | thanks guys | 15:06 |
mriedem | dansmith: given tempest i'm sure it is right after the instance goes active when they try to hit the floating ips api | 15:06 |
mriedem | and at that point nw info_cache is empty | 15:07 |
JayF | https://review.openstack.org/#/c/176395/ has 10 +1s, fixes a major bug with Ironic<>Nova + Configdrive. If someone could core review it I'd be appreciative | 15:07 |
dansmith | and with nova-net, huh? | 15:07 |
*** apuimedo has quit IRC | 15:07 | |
*** salv-orlando has quit IRC | 15:08 | |
*** annegentle has joined #openstack-nova | 15:09 | |
*** tonytan4ever has joined #openstack-nova | 15:09 | |
dansmith | mriedem: so was this the original ssh timeout bug that we thought was only neutron? | 15:09 |
openstackgerrit | Sylvain Bauza proposed openstack/nova: WIP: Create RequestSpec object https://review.openstack.org/145528 | 15:09 |
*** mwagner_lap has quit IRC | 15:10 | |
dansmith | mriedem: like, I didn't think we saw this 400 error every time we get that timeout | 15:10 |
*** apmelton_away is now known as apmelton | 15:10 | |
mriedem | dansmith: bug 1249065 is unrelated to the ssh timeout bug(s) | 15:10 |
openstack | bug 1249065 in OpenStack Compute (nova) "Nova throws 400 when attempting to add floating ip (No nw_info cache associated with instance)" [Medium,In progress] https://launchpad.net/bugs/1249065 - Assigned to Matt Riedemann (mriedem) | 15:10 |
*** wwriverrat has quit IRC | 15:10 | |
mriedem | well, unrelated in that the tests aren't trying to do ssh thingies | 15:10 |
dansmith | mriedem: https://bugs.launchpad.net/nova/+bug/1249065/comments/1 | 15:11 |
dansmith | so, I thought we only did async network allocation on neutron, but I guess we do it even for nova-net | 15:12 |
mriedem | yeah | 15:13 |
mriedem | dansmith: i was thinking more like this http://logs.openstack.org/04/173304/7/check/check-tempest-dsvm-nova-v21-full/69342e7/console.html.gz#_2015-04-30_08_48_29_781 | 15:13 |
BadCub | Hiya folks. Would it be possible to get some core reviews on: https://review.openstack.org/#/c/176395/ ? | 15:13 |
mriedem | the tempest api tests aren't doing ssh like the scenario tests | 15:13 |
*** annegentle has quit IRC | 15:14 | |
mriedem | if bug 1249065 only ever shows up in that rescue test, might be a problem with rescue | 15:14 |
openstack | bug 1249065 in OpenStack Compute (nova) "Nova throws 400 when attempting to add floating ip (No nw_info cache associated with instance)" [Medium,In progress] https://launchpad.net/bugs/1249065 - Assigned to Matt Riedemann (mriedem) | 15:14 |
*** annegentle has joined #openstack-nova | 15:14 | |
lucasagomes | hi all, if you have a time please take a look at this patch for Ironic https://review.openstack.org/#/c/176395/ it's quite small (+3,-2LOC) and fix a bug with the configdrive in Ironic | 15:14 |
lucasagomes | thanks | 15:14 |
mriedem | who wtf http://logs.openstack.org/32/178332/5/check/check-tempest-dsvm-postgres-full/fb52792/logs/screen-n-api.txt.gz?level=TRACE#_2015-05-06_16_43_36_061 | 15:15 |
mriedem | i've never seen that before ^ | 15:15 |
mriedem | oh negative test, so hitting a db api thing in postgresql | 15:16 |
dansmith | mriedem: unfortunately that doesn't generate any logs on the api or compute side | 15:16 |
dansmith | the 400 for associate that is | 15:16 |
dansmith | and on rescue we don't have to do an allocation step, AFAIK | 15:17 |
mriedem | dansmith: same http://logs.openstack.org/04/173304/7/check/check-tempest-dsvm-nova-v21-full/69342e7/console.html.gz#_2015-04-30_09_04_21_525 | 15:17 |
dansmith | right | 15:17 |
dansmith | that's the one I'm talking about | 15:17 |
mriedem | well in the api log it's just info level http://logs.openstack.org/04/173304/7/check/check-tempest-dsvm-nova-v21-full/69342e7/logs/screen-n-api.txt.gz#_2015-04-30_08_48_27_300 | 15:18 |
mriedem | but it's the same, and not helpful | 15:18 |
mriedem | wonder, could we just add a check somewhere in the network api layer such that if instance.info_cache goes from something to nothing, we log a warning? | 15:19 |
mriedem | like in the refresh decorator? | 15:19 |
dansmith | well, in the instance object would catch everything I think | 15:19 |
mriedem | sure, anything | 15:19 |
mriedem | just, once you have some nw info, is it valid to not anymore unless you're deleting or something? | 15:19 |
dansmith | yeah | 15:20 |
dansmith | we don't even save the info cache if it's None | 15:20 |
dansmith | on instance.save() | 15:20 |
alaski | mriedem: just if you deallocate I think | 15:21 |
*** dsanders has joined #openstack-nova | 15:22 | |
dansmith | alaski: what? | 15:22 |
alaski | for nw_info to go from something to nothing | 15:22 |
alaski | if you deallocate the networks it would | 15:22 |
dansmith | which you do when? | 15:23 |
dansmith | besides delete | 15:23 |
alaski | only delete afaik, but I am only familiar with simple use cases | 15:23 |
dansmith | what about migration? | 15:23 |
dansmith | don't we do some deallocation on the old node on confirm? | 15:24 |
dansmith | obviously it shouldn't null out info_cache, but.. | 15:24 |
*** Nikolay_St has quit IRC | 15:24 | |
mriedem | i'm only seeing deallocate_network on terminate and build failed/reschedule | 15:25 |
alaski | dansmith: it's setup_networks_on_host in that case, with teardown=True. not sure what that actually does | 15:25 |
dansmith | I'm only seeing info_cache.delete() in delete_instance and local_delete | 15:26 |
dansmith | alaski: oh, right, right | 15:26 |
dansmith | setup(teardown=True) <--- terrible | 15:26 |
alaski | yep | 15:26 |
dansmith | mriedem: that call is _add_floating_ip() in floating_ips.py right? | 15:28 |
mriedem | dansmith: yeah https://review.openstack.org/#/c/171391/3/nova/api/openstack/compute/contrib/floating_ips.py | 15:29 |
dansmith | so, that error is kinda misleading | 15:30 |
dansmith | I was thinking that it meant there was no record in the db | 15:30 |
openstackgerrit | dave-mcnally proposed openstack/nova: Add policy to cover snapshotting of volume backed instances https://review.openstack.org/176805 | 15:31 |
dansmith | but it actually could mean that it's there and empty | 15:31 |
mriedem | dansmith: right | 15:31 |
*** browne has quit IRC | 15:31 | |
dansmith | because of that completely useless helper | 15:31 |
*** oro has quit IRC | 15:31 | |
mriedem | wtf, nova.tests.unit.test_crypto.KeyPairTest.test_generate_key_pair times out for me locally every time i run it | 15:32 |
dansmith | that change you have proposed is kinda icky because it could generate a ton of back and forth with neutron | 15:33 |
mriedem | dansmith: well, i originally proposed it b/c ganso in here was saying he was recreating the test failure all of the time on his juno ci setup, | 15:33 |
mriedem | so wipped that up for him to see if it did anything, | 15:33 |
dansmith | okay | 15:33 |
mriedem | then he changed something and wasn't hitting the problem anymore | 15:33 |
*** shiva has joined #openstack-nova | 15:35 | |
*** Marga_ has joined #openstack-nova | 15:35 | |
*** marun has joined #openstack-nova | 15:35 | |
shiva | Hi | 15:35 |
*** dsanders has quit IRC | 15:35 | |
*** Marga_ has quit IRC | 15:36 | |
shiva | Wanted to know if I can get some help on VMWare + NOVA | 15:36 |
shiva | especially on VM placement stuff with NOVA on a cluster | 15:36 |
*** Marga_ has joined #openstack-nova | 15:36 | |
openstackgerrit | Dan Smith proposed openstack/nova: Log the actual instance.info_cache when empty in floating ip associate https://review.openstack.org/181035 | 15:36 |
dansmith | mriedem: ^ that will give us some info about what is actually in info_cache | 15:37 |
cfriesen | is there a way to run the tox pep8 tests only on specific files? | 15:37 |
dansmith | not sure if that can be warning, but seems appropriate given what that means | 15:37 |
mriedem | dansmith: yeah, and assuming we can recreate it on that change | 15:37 |
mriedem | out of thousands | 15:37 |
dansmith | mriedem: well, we can just merge that and wait | 15:37 |
mriedem | dansmith: you need it on v2 then also | 15:37 |
shiva | I want to see if there is a way from nova to query which ESX host is the VM Placed on when deployed | 15:38 |
openstackgerrit | Chris Friesen proposed openstack/nova: unify libvirt driver checks for qemu https://review.openstack.org/175542 | 15:38 |
mriedem | cfriesen: nope | 15:38 |
mriedem | shiva: see topic please | 15:38 |
mriedem | shiva: you probably want to find garyk | 15:39 |
dansmith | mriedem: okay | 15:39 |
dansmith | shiva: try #openstack-vmware | 15:39 |
openstackgerrit | Dan Smith proposed openstack/nova: Log the actual instance.info_cache when empty in floating ip associate https://review.openstack.org/181035 | 15:39 |
shiva | thanks dansmith | 15:39 |
openstackgerrit | Matt Riedemann proposed openstack/nova: Log the actual instance.info_cache when empty in floating ip associate https://review.openstack.org/181035 | 15:40 |
*** nelsnelson has joined #openstack-nova | 15:40 | |
*** sdake has quit IRC | 15:44 | |
*** sdake has joined #openstack-nova | 15:44 | |
*** zhenguo has quit IRC | 15:44 | |
*** shiva has left #openstack-nova | 15:47 | |
*** lpetrut has quit IRC | 15:48 | |
*** artom has joined #openstack-nova | 15:49 | |
*** markus_z has quit IRC | 15:49 | |
*** kylek3h has joined #openstack-nova | 15:49 | |
garyk | dansmith: no one really uses #openstack-vmware any more - we prefer to do it here | 15:49 |
dansmith | garyk: okay, well, still, non-dev discussion belongs elsewhere | 15:50 |
*** mtanino has joined #openstack-nova | 15:50 | |
garyk | dansmith: sure, but how areyou sure that it is not dev related | 15:51 |
*** artom has quit IRC | 15:51 | |
*** artom has joined #openstack-nova | 15:51 | |
dansmith | garyk: seems pretty clear to me | 15:52 |
dansmith | assume shiva would have told us otherwise when we pointed to topic | 15:52 |
*** mtanino_ has joined #openstack-nova | 15:52 | |
*** matrohon has quit IRC | 15:53 | |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: Add functional test for server group https://review.openstack.org/147514 | 15:54 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: scheduler:refactor group info in filter_properties https://review.openstack.org/148275 | 15:54 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: api: add soft-affinity policies for server groups https://review.openstack.org/147520 | 15:54 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: Add functional tests to soft affinity capabilities https://review.openstack.org/147521 | 15:55 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: scheduler: add num of instance per host in a group https://review.openstack.org/148277 | 15:55 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: object:add count_members_by_hosts to InstanceGroup https://review.openstack.org/147517 | 15:55 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: scheduler: add soft-(anti-)affinity weighers https://review.openstack.org/147516 | 15:55 |
*** mtanino has quit IRC | 15:55 | |
garyk | dansmith: it was actually related - we have a patch in review for donkeys years (parity to a libvirt one added in J). that would address the question | 15:56 |
dansmith | asking a question about a missing feature you have a patch up for is not dev-related, IMHO, but whatever :) | 15:56 |
*** angdraug has joined #openstack-nova | 15:57 | |
garyk | dansmith: helpful | 15:57 |
*** dsanders has joined #openstack-nova | 15:58 | |
*** mtanino_ has quit IRC | 15:59 | |
*** mwagner_lap has joined #openstack-nova | 16:00 | |
*** moshele has quit IRC | 16:01 | |
*** jlanoux has quit IRC | 16:02 | |
*** dboik_ has quit IRC | 16:04 | |
*** bkopilov has quit IRC | 16:04 | |
*** dboik has joined #openstack-nova | 16:04 | |
*** otter768 has joined #openstack-nova | 16:07 | |
*** mpavone has quit IRC | 16:07 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 16:08 |
*** bkopilov has joined #openstack-nova | 16:09 | |
*** salv-orlando has joined #openstack-nova | 16:09 | |
*** sahid has quit IRC | 16:10 | |
*** otter768 has quit IRC | 16:11 | |
*** salv-orlando has quit IRC | 16:12 | |
*** browne has joined #openstack-nova | 16:13 | |
openstackgerrit | Michal Rostecki proposed openstack/nova-specs: Dedicate aggregates for specific tenants https://review.openstack.org/181054 | 16:13 |
*** pask81 has quit IRC | 16:14 | |
*** lpetrut has joined #openstack-nova | 16:15 | |
*** unicell has quit IRC | 16:17 | |
*** derekh has quit IRC | 16:19 | |
*** gokrokve_ has joined #openstack-nova | 16:22 | |
*** changbl has joined #openstack-nova | 16:23 | |
*** gokrokve_ has quit IRC | 16:23 | |
*** gokrokve_ has joined #openstack-nova | 16:24 | |
*** Marga_ has quit IRC | 16:24 | |
*** dikonoor has quit IRC | 16:24 | |
*** mtanino has joined #openstack-nova | 16:25 | |
*** gokrokve has quit IRC | 16:25 | |
garyk | dansmith: please see comments on https://review.openstack.org/#/c/181035/ | 16:26 |
dansmith | garyk: and it matters for a single param why? | 16:27 |
garyk | i think that you misunderstood. | 16:28 |
dansmith | the few things above it that I see with single format variables all use % | 16:28 |
dansmith | so I think I'm good | 16:28 |
mriedem | it's the delayed format processing with log.warning that garyk is probably referring to | 16:29 |
mriedem | with , vs % | 16:29 |
garyk | sure, in the past i have been asked to do the , instead of the % - so that the assignment is not done in the log module. but whatever you choose | 16:29 |
*** oro has joined #openstack-nova | 16:29 | |
*** shakamunyi has quit IRC | 16:29 | |
garyk | mriedem: yes, that is what i am referring to. | 16:29 |
dansmith | mriedem: I don't get what the difference is for this though | 16:30 |
mriedem | in the case that someone isn't logging at warning+ level, which i'd say is bonkers if they aren't | 16:30 |
dansmith | in that case what? | 16:30 |
mriedem | that the substitution doesn't need to be performed if you're not logging at warning level | 16:31 |
*** Marga_ has joined #openstack-nova | 16:31 | |
dansmith | still not getting why it matters, unless you're saying it saves the few cycles of doing a thing that doesn't need to be done | 16:32 |
mriedem | theoretically that's the point | 16:32 |
mriedem | but i'm not -1 on this :) | 16:32 |
garyk | nor am i i just thought that dan would be open to a suggestion... | 16:32 |
*** xyang1 has quit IRC | 16:32 | |
*** Murali has joined #openstack-nova | 16:32 | |
*** lucasagomes has quit IRC | 16:33 | |
dansmith | heh, okay | 16:33 |
dansmith | it's only in a rare error case, so not like we're doing it all the time either | 16:33 |
mriedem | http://docs.openstack.org/developer/oslo.i18n/guidelines.html#adding-variables-to-log-messages | 16:33 |
mriedem | officially ^ | 16:33 |
dansmith | okay, if I have to rebase I'll change it | 16:33 |
garyk | thanks | 16:34 |
*** tpeoples has quit IRC | 16:34 | |
*** emagana has quit IRC | 16:35 | |
*** Murali_ has joined #openstack-nova | 16:35 | |
*** alanf-mc has joined #openstack-nova | 16:36 | |
*** Murali__ has joined #openstack-nova | 16:36 | |
openstackgerrit | Michal Rostecki proposed openstack/nova-specs: Dedicate aggregates for specific tenants https://review.openstack.org/181054 | 16:37 |
*** Murali has quit IRC | 16:37 | |
*** Murali__ is now known as Murali | 16:37 | |
*** annegentle has quit IRC | 16:39 | |
*** Murali_ has quit IRC | 16:40 | |
*** rushiagr_away is now known as rushiagr | 16:40 | |
*** emagana has joined #openstack-nova | 16:42 | |
*** Nic has joined #openstack-nova | 16:43 | |
*** gszasz has joined #openstack-nova | 16:43 | |
*** alanf-mc has quit IRC | 16:43 | |
*** tsekiyama has joined #openstack-nova | 16:44 | |
*** alanf-mc has joined #openstack-nova | 16:45 | |
*** dulek has quit IRC | 16:45 | |
*** patrickeast has joined #openstack-nova | 16:46 | |
*** IanGovett1 has quit IRC | 16:46 | |
*** unicell has joined #openstack-nova | 16:47 | |
*** romainh has left #openstack-nova | 16:47 | |
*** IanGovett has joined #openstack-nova | 16:48 | |
*** Murali has quit IRC | 16:49 | |
*** neelashah has quit IRC | 16:50 | |
sdague | dansmith: so http://logs.openstack.org/22/162722/7/gate/gate-grenade-dsvm/b2f3148/logs/old/screen-n-cpu.txt.gz ... that's one of those fails where the server worked fine, then the check once the services went down failed | 16:50 |
dansmith | and that's not mine? | 16:51 |
* dansmith checks | 16:51 | |
dansmith | okay, well, that makes me feel better that it's not my problem | 16:51 |
sdague | yeh, I don't think that patch set of yours is the issue | 16:51 |
dansmith | well, same issue in two runs, I was worried, especially given the change I was making | 16:52 |
dansmith | but cool | 16:52 |
*** jichen has quit IRC | 16:52 | |
*** sdake_ has joined #openstack-nova | 16:52 | |
dansmith | sdague: so is this spiking? | 16:52 |
openstackgerrit | Gorka Eguileor proposed openstack/nova: Preserve usage and reservations on quota deletion https://review.openstack.org/163423 | 16:52 |
sdague | dansmith: I'm not sure it's spiking, but I'm trying to get ahead of this | 16:52 |
sdague | because it's weird | 16:52 |
dansmith | well, | 16:52 |
dansmith | what I mean is.. this appears to be a thing that just started happening at a decent rate? | 16:53 |
dansmith | because twice for me is pretty good odds | 16:53 |
sdague | and... this particular instance seems to be in the middle of the heal_instance_info_cache when the compute got a SIGTERM | 16:53 |
*** ildikov has quit IRC | 16:53 | |
*** garyk has quit IRC | 16:53 | |
sdague | http://logs.openstack.org/22/162722/7/gate/gate-grenade-dsvm/b2f3148/logs/old/screen-n-cpu.txt.gz#_2015-05-07_04_12_20_488 | 16:53 |
*** vilobhmm1 has joined #openstack-nova | 16:54 | |
dansmith | sdague: but that's the end of the new run, right? | 16:54 |
sdague | that's the end of the old | 16:54 |
dansmith | oh, this isn't partial | 16:54 |
sdague | the explosition is testing the networking of the computes when all the services are down | 16:54 |
sdague | correct | 16:54 |
*** pkoniszewski has quit IRC | 16:54 | |
dansmith | explosition? awesome | 16:54 |
*** claudiub has joined #openstack-nova | 16:54 | |
mriedem | http://goo.gl/Kza2rc | 16:55 |
dansmith | so wait, you're saying we kill compute to do the upgrade and you think that means we're suddenly unable to ping the guests that should still be running? | 16:55 |
dansmith | sdague: & | 16:55 |
dansmith | er, sdague ^ | 16:55 |
*** sdake has quit IRC | 16:56 | |
dansmith | mriedem: I think that's too generic of a thing to query for | 16:56 |
dansmith | sdague: do you have a link to that service signal change we made recently? | 16:56 |
sdague | dansmith: I think some times when we take the services down we lose working guests | 16:57 |
*** tjones2 has left #openstack-nova | 16:57 | |
sdague | well, I know that some times when we take services down we lose working guests, I think it may be related to partially finished work they have | 16:57 |
dansmith | you think that because we pinged them earlier, | 16:57 |
sdague | dansmith: well I have a link to the revert which I pushed | 16:57 |
dansmith | but that doesn't mean that they didn't die before the service went doen | 16:57 |
dansmith | sdague: revert? | 16:58 |
sdague | https://review.openstack.org/#/c/180964/ | 16:58 |
dansmith | I got the original one | 16:58 |
dansmith | ah, I meant this: https://review.openstack.org/#/c/176777/ | 16:58 |
dansmith | all the things we do for networking are spawning processes | 16:58 |
sdague | dansmith: ok, sure, but that code isn't really functioning yet, because we're double killing everything | 16:58 |
dansmith | so I wanted to look at this to see if we messed up SIGCHLD somehow | 16:59 |
dansmith | and are taking out dnsmasq when we exit or something | 16:59 |
sdague | ok, that's fair | 16:59 |
dansmith | but I guess we're not changing SIGCHLD behavior, just what we do when we term _as_ a child | 16:59 |
mriedem | zzzeek: this has to be 293 now https://review.openstack.org/#/c/154243/ | 17:00 |
*** exploreshaifali has joined #openstack-nova | 17:00 | |
*** emagana has quit IRC | 17:00 | |
sdague | so... the number of things done between the working pings, and the failing pings are actually quite small, as this happens post tempest runs | 17:00 |
mriedem | zzzeek: ping me when it's up and i'll +2 | 17:00 |
*** Marga_ has quit IRC | 17:00 | |
zzzeek | mriedem: heh no pressure right :) | 17:00 |
zzzeek | quick before its 294! | 17:01 |
mriedem | right :) | 17:01 |
*** emagana has joined #openstack-nova | 17:01 | |
dansmith | sdague: timing being like a second or what? | 17:02 |
sdague | dansmith: sorry, timing of which? | 17:03 |
dansmith | sdague: between ping and fail | 17:05 |
* dansmith is doing like eight things right now3 | 17:05 | |
sdague | dansmith: ok, go do other things, we can circle back | 17:07 |
dansmith | well, this seems important | 17:07 |
dansmith | sdague: do you know what order services come down in? | 17:08 |
dansmith | wait, this is nova-net right? | 17:08 |
sdague | it's nova net | 17:09 |
sdague | I can look them up, conductor is definitely down before compute here, based on the sigterm lines | 17:10 |
sdague | I'm sort of wondering if the double kill is exacerbating things, because services don't get a chance to finish what they are doing. | 17:10 |
sdague | which is why I proposed that revert to devstack so we can do this the way we're supposed to be | 17:11 |
dansmith | well, the double kill of conductor wouldn't affect this task in that way | 17:12 |
dansmith | because this task makes several calls, each of which can straddle the kill | 17:12 |
dansmith | even gracefully | 17:12 |
dansmith | I don't know what we do for periodics though, let me look | 17:12 |
dansmith | and I think conductor wasn't doing graceful shutdown anyway, IIRC | 17:12 |
*** sdake has joined #openstack-nova | 17:13 | |
*** yamahata has quit IRC | 17:14 | |
*** tonytan4ever has quit IRC | 17:14 | |
*** tonytan4ever has joined #openstack-nova | 17:16 | |
*** jistr has quit IRC | 17:17 | |
*** sdake_ has quit IRC | 17:17 | |
dansmith | sdague: so, when we get that signal in compute, we stop the roc server, which I think cuts us off from any replies we were going to get | 17:18 |
*** lpetrut has quit IRC | 17:18 | |
dansmith | er, rpc server | 17:18 |
dansmith | and then waits for any threads currently running to finish | 17:18 |
dansmith | but it doesn't wait for periodics | 17:18 |
*** IanGovett has quit IRC | 17:19 | |
dansmith | so I don't *think* that the double kill is making it worse that we exit in the middle of that periodic | 17:19 |
sdague | right, so the compute is definitely hanging in heal_info_cache when this particular fail happens | 17:19 |
sdague | which may or may not be relevant | 17:19 |
*** VW_ has quit IRC | 17:19 | |
sdague | normally it would have completed | 17:19 |
dansmith | I also don't think there is anything in that periodic which can cause the networks to actually go down if we fail in the middle | 17:19 |
openstackgerrit | Daniel Berrange proposed openstack/nova: objects: define the ImageMeta & ImageMetaProps objects https://review.openstack.org/76234 | 17:20 |
openstackgerrit | Daniel Berrange proposed openstack/nova: objects: add a FlexibleBoolean field type https://review.openstack.org/181085 | 17:20 |
openstackgerrit | Daniel Berrange proposed openstack/nova: objects: treat EnumField and ListOfEnumField classes as abstract https://review.openstack.org/181086 | 17:20 |
sdague | however with conductor down, that 3rd line is it going out to get the info from conductor, and it's hung there | 17:20 |
sdague | dansmith: we don't ever try to do things to clean up security groups if hosts are lost? | 17:20 |
dansmith | sdague: which third line? | 17:20 |
dansmith | sdague: what do you mean "if hosts are lost" ? | 17:21 |
*** VW_ has joined #openstack-nova | 17:21 | |
sdague | sorry, now I'm doing 2 things. :) Let me dig up | 17:22 |
sdague | http://logs.openstack.org/22/162722/7/gate/gate-grenade-dsvm/b2f3148/logs/old/screen-n-cpu.txt.gz#_2015-05-07_04_12_20_514 | 17:22 |
openstackgerrit | Matt Riedemann proposed openstack/nova: Show 'reserved' status in os-fixed-ips https://review.openstack.org/168966 | 17:22 |
sdague | if you look at the heal before that - http://logs.openstack.org/22/162722/7/gate/gate-grenade-dsvm/b2f3148/logs/old/screen-n-cpu.txt.gz#_2015-05-07_04_11_20_488 | 17:22 |
dansmith | sdague: actually probably not going to conductor for that | 17:23 |
dansmith | sdague: in the flavor case, that's probably just processing system_metadata | 17:23 |
*** IanGovett has joined #openstack-nova | 17:23 | |
dansmith | although it needs to go to conductor to get extra_specs I guess | 17:23 |
*** gszasz has quit IRC | 17:23 | |
dansmith | either way, that task is full of calls to conductor, so if it's not that lazy-load it's something else for sure | 17:23 |
dansmith | sdague: on the secgroups thing, were you asking if we nuke secgroups if we think the instance has gone away because of a failure to refresh the info cache? | 17:24 |
*** armax has joined #openstack-nova | 17:25 | |
*** Marga_ has joined #openstack-nova | 17:25 | |
sdague | or something else tickled our iptables stack | 17:26 |
sdague | but yes | 17:26 |
*** xyang1 has joined #openstack-nova | 17:26 | |
dansmith | sdague: shall we dump iptables after this failure, or do we do that already? | 17:26 |
sdague | so... that patch is inbound | 17:27 |
openstackgerrit | Daniel Berrange proposed openstack/nova: objects: add a FlexibleBoolean field type https://review.openstack.org/181085 | 17:27 |
openstackgerrit | Daniel Berrange proposed openstack/nova: objects: treat EnumField class as abstract https://review.openstack.org/181086 | 17:27 |
openstackgerrit | Daniel Berrange proposed openstack/nova: objects: define the ImageMeta & ImageMetaProps objects https://review.openstack.org/76234 | 17:27 |
sdague | let me see if it landed yet | 17:27 |
dansmith | sdague: I don't think there's anything in there that would be triggering us doing that, FWIW, and I think we'd see a bunch of log noise if so | 17:27 |
dansmith | sdague: okay cool | 17:27 |
*** oro has quit IRC | 17:27 | |
sdague | yeh, I'd have thought so as well | 17:27 |
*** oro has joined #openstack-nova | 17:28 | |
openstackgerrit | Eric Brown proposed openstack/nova: VMware: update supported vsphere 6.0 os types https://review.openstack.org/181020 | 17:29 |
sdague | however, the fact that the guest pinged here - http://logs.openstack.org/22/162722/7/gate/gate-grenade-dsvm/b2f3148/logs/grenade.sh.txt.gz#_2015-05-07_04_12_11_959 | 17:29 |
sdague | then failed here - http://logs.openstack.org/22/162722/7/gate/gate-grenade-dsvm/b2f3148/logs/grenade.sh.txt.gz#_2015-05-07_04_12_43_954 | 17:29 |
dansmith | sdague: we also don't see any traces from some call raising an exception due to rpc being down or anything | 17:29 |
dansmith | we don't have any except Exception: statements in there | 17:29 |
sdague | dansmith: sure, but we wouldn't | 17:30 |
sdague | that timeout is 60 seconds or something | 17:30 |
dansmith | sdague: dude, that's 30 whole seconds! | 17:30 |
*** Marga_ has quit IRC | 17:30 | |
sdague | and all the services go down | 17:30 |
sdague | in a couple of seconds | 17:30 |
dansmith | sdague: right, what I'm saying is: | 17:30 |
*** Marga_ has joined #openstack-nova | 17:30 | |
dansmith | sdague: if taking down conductor first caused us to receive an exception and we were going to turn that into "screw with security groups because instance is fubar" we'd have seen the trace that makes it think that's a good idea | 17:31 |
sdague | dansmith: ok sure | 17:31 |
*** Marga_ has quit IRC | 17:31 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 17:31 |
sdague | so the only other active thing that happened between that guest working, and then failing later, is we build the cinder boot from volume guest | 17:32 |
*** danpb has quit IRC | 17:32 | |
dansmith | ah | 17:32 |
dansmith | which does screw with secgroups | 17:32 |
sdague | which means this race might entirely be related to us self racing on secgroups | 17:32 |
sdague | yep | 17:32 |
dansmith | I think firewalls are overrated anyway, we should just remove that code | 17:33 |
dansmith | sdague: did we ever find any damning evidence in iptables dumps to support the races we still think are in that code? | 17:33 |
dansmith | meaning, | 17:33 |
dansmith | unless we see some obviously stupid rule in the forthcoming dump, I wonder if we'll be even able to tell what's going on | 17:34 |
dansmith | also | 17:34 |
sdague | we haven't yet, I'm pretty close to dumping iptables before and after every guest boot here so we can walk through it on working and non working | 17:34 |
dansmith | sdague: that is trying to ping $subnet.1 .. is that really the ip of the instance? | 17:34 |
dansmith | sdague: +1 | 17:34 |
*** mspreitz has quit IRC | 17:34 | |
sdague | dansmith: yes, that's what nova gave for the floating ip | 17:35 |
dansmith | hmm | 17:35 |
*** Marga_ has joined #openstack-nova | 17:35 | |
*** tjones1 has joined #openstack-nova | 17:35 | |
sdague | http://logs.openstack.org/22/162722/7/gate/gate-grenade-dsvm/b2f3148/logs/grenade.sh.txt.gz#_2015-05-07_04_11_21_836 | 17:35 |
*** tjones1 has quit IRC | 17:36 | |
dansmith | yeah, just reading | 17:36 |
dansmith | there is nothing in compute cleanup_host() that mucks with secgroups, and libvirt doesn't implement that cleanup hook either | 17:38 |
*** dsanders has quit IRC | 17:39 | |
*** yamahata has joined #openstack-nova | 17:40 | |
*** annegentle has joined #openstack-nova | 17:40 | |
*** emagana has quit IRC | 17:40 | |
dansmith | sdague: so, what about dumping the guest console log to see if something we did with one of the other services caused it heartache? | 17:41 |
dansmith | are these BFV guests by any chance? | 17:41 |
*** isd has joined #openstack-nova | 17:42 | |
*** erlon has joined #openstack-nova | 17:43 | |
*** achanda has joined #openstack-nova | 17:43 | |
*** emagana has joined #openstack-nova | 17:43 | |
sdague | dansmith: the 2nd one is | 17:44 |
sdague | but that one is working | 17:44 |
dansmith | heh | 17:44 |
sdague | in this job | 17:44 |
sdague | in your original failure, the BFV was the on failing | 17:44 |
dansmith | interesting | 17:44 |
dansmith | that does kinda sound like a secgroups thing | 17:44 |
*** Nikolay_St has joined #openstack-nova | 17:45 | |
*** MVenesio has joined #openstack-nova | 17:45 | |
sdague | dansmith: so... how would you expect to dump the console all the services are down | 17:45 |
*** yassine_ has quit IRC | 17:45 | |
dansmith | and I know it would please you to no end for it to be secgroups race -related :) | 17:45 |
dansmith | sdague: it's just a file on disk, right? | 17:45 |
*** anteaya has quit IRC | 17:45 | |
dansmith | sdague: we haven't fixed that grows-forever-muahaha bug yet, AFAIK | 17:45 |
*** tonytan4ever has quit IRC | 17:45 | |
sdague | oh, good point, just a find ... ? | 17:46 |
sdague | you remember where they are | 17:46 |
*** annegentle has quit IRC | 17:46 | |
dansmith | I can look, jsut a sec | 17:46 |
dansmith | sdague: /opt/stack/data/nova/instances/c5967308-4cf4-4114-853e-dd6bfcd64aa1/console.log | 17:47 |
dansmith | should be easy to snag, maybe even just snag the one that we can't ping? | 17:47 |
sdague | ok, I'll code up something to capture all those and put them in the dump file, that will happen later today | 17:48 |
dansmith | cool | 17:48 |
*** emagana has quit IRC | 17:48 | |
dansmith | while you do that, I'll try to find other employment so I don't have to worry about this anymore... :P | 17:48 |
*** kaisers has quit IRC | 17:48 | |
sdague | heh | 17:49 |
*** jistr has joined #openstack-nova | 17:49 | |
*** kaisers has joined #openstack-nova | 17:49 | |
*** Nikolay_St has quit IRC | 17:50 | |
*** claudiub has quit IRC | 17:53 | |
*** gokrokve has joined #openstack-nova | 17:53 | |
vilobhmm1 | dims : ping | 17:54 |
dims | vilobhmm1: pong | 17:54 |
vilobhmm1 | johnthetubaguy, dims : can you please review https://review.openstack.org/#/c/138607/ have addressed all the concerns raised and also updated the migration plan | 17:55 |
*** tonytan4ever has joined #openstack-nova | 17:55 | |
dims | vilobhmm1: queuing it up | 17:55 |
*** gokrokve_ has quit IRC | 17:57 | |
vilobhmm1 | ok thanks | 17:57 |
*** gokrokve has quit IRC | 17:59 | |
*** derekh has joined #openstack-nova | 17:59 | |
*** rushiagr is now known as rushiagr_away | 18:00 | |
*** tzumainn_ has joined #openstack-nova | 18:00 | |
*** gokrokve has joined #openstack-nova | 18:03 | |
*** vilobhmm1 has quit IRC | 18:03 | |
*** vilobhmm1 has joined #openstack-nova | 18:03 | |
*** moshele has joined #openstack-nova | 18:04 | |
*** vilobhmm1 has quit IRC | 18:04 | |
*** vilobhmm1 has joined #openstack-nova | 18:04 | |
Nic | So, we found this in integration testing: https://bugs.launchpad.net/nova/+bug/1452840 | 18:04 |
openstack | Launchpad bug 1452840 in OpenStack Compute (nova) "libvirt: nova's detach_volume silently fails sometimes" [Undecided,New] | 18:04 |
Nic | I've come up with an extraordinarily hackish workaround that involves detecting that things have gone wrong, and shelling out to virsh | 18:05 |
*** emagana has joined #openstack-nova | 18:05 | |
Nic | I suspect that this approach is a non-starter | 18:05 |
*** ericksonsantos has quit IRC | 18:05 | |
*** ildikov has joined #openstack-nova | 18:05 | |
dansmith | yeah | 18:05 |
*** tellesnobrega has quit IRC | 18:05 | |
*** raildo has quit IRC | 18:06 | |
*** htruta has quit IRC | 18:06 | |
Nic | But I figure folks with more insight than me could lend some advice | 18:06 |
dansmith | what does shelling out to virsh help? are you doing something else or just trying the same thing with virsh? | 18:06 |
*** afaranha has quit IRC | 18:06 | |
*** annegentle has joined #openstack-nova | 18:06 | |
Nic | It's all very black-box, since Nova fires a libvirt method which disappears down a shared-object rabbit-hole | 18:07 |
Nic | But, there is a disconnect between what QEMU and libvirt think is going on with the guest | 18:07 |
mriedem | Nic: so libvirt is saying it's detached but it's not? | 18:07 |
*** otter768 has joined #openstack-nova | 18:07 | |
dansmith | Nic: meaning nova has a cached notion of the guest and virsh builds it from scratch, and thus does the right thing? | 18:08 |
Nic | No, libvirt is saying it's detaching it, but failing, and QEMU says on subsequent detach requests that the device is gone (but it's not) | 18:08 |
*** sdake has quit IRC | 18:08 | |
Nic | I see no evidence of any explicit caching going on, but there is some manner of disconnect happening | 18:09 |
dansmith | well, | 18:09 |
dansmith | call it caching or "state coherence" or something | 18:09 |
*** sdake has joined #openstack-nova | 18:09 | |
Nic | If you ask virsh to smash the device, it works fine | 18:09 |
sdague | ok, I'm going to get some food, then will ponder the bigger debug dumping for this | 18:09 |
dansmith | sdague: cool | 18:10 |
dansmith | Nic: I think we should try to get danpb to look at that bug | 18:10 |
Nic | I'm 99.999% sure it's nothing particular to Nova | 18:10 |
mriedem | Nic: can you reproduce with newer libvirt, like 1.2.9? | 18:10 |
dansmith | Nic: right, so danpb is the right person to know if it's a libvirt thing and fix it :) | 18:10 |
dansmith | Nic: or the right person to know if it's a usage issue or anything else | 18:11 |
Nic | I just grabbed a stock Trusty image and devstack off of master, and it's still b0rken there | 18:11 |
Nic | Got a preferred PPA for bleeding-edge libvirt? | 18:12 |
*** otter768 has quit IRC | 18:12 | |
mriedem | Nic: try 1.2.9 b/c that's what fc21 has and we have a job to test that | 18:12 |
mriedem | otherwise we're running/gating on trusty with libvirt 1.2.2, same as what you're reporting here | 18:13 |
*** VW_ has quit IRC | 18:13 | |
*** vilobhmm1 has quit IRC | 18:13 | |
*** gokrokve has quit IRC | 18:13 | |
*** gokrokve has joined #openstack-nova | 18:14 | |
*** tjones1 has joined #openstack-nova | 18:15 | |
*** VW_ has joined #openstack-nova | 18:16 | |
*** nikhil_k_ is now known as nikhil_k-afk | 18:17 | |
*** ildikov has quit IRC | 18:17 | |
*** baoli has quit IRC | 18:17 | |
*** baoli has joined #openstack-nova | 18:18 | |
*** baoli has quit IRC | 18:18 | |
*** vilobhmm1 has joined #openstack-nova | 18:18 | |
*** derekh has quit IRC | 18:18 | |
*** gokrokve has quit IRC | 18:19 | |
*** anteaya has joined #openstack-nova | 18:19 | |
*** VW__ has joined #openstack-nova | 18:19 | |
*** VW_ has quit IRC | 18:21 | |
*** sudipto has joined #openstack-nova | 18:21 | |
mriedem | i would like to see a few more +1s on this https://review.openstack.org/#/c/176395/ | 18:21 |
mriedem | JayF: can you make that happen ^ ? :) | 18:21 |
JayF | -.- | 18:21 |
*** vilobhmm1 has quit IRC | 18:21 | |
*** tellesnobrega has joined #openstack-nova | 18:22 | |
JayF | mriedem: https://review.openstack.org/#/c/180632/1 there's a small followup patch too | 18:22 |
mriedem | reading it now | 18:22 |
*** vilobhmm1 has joined #openstack-nova | 18:22 | |
JayF | tyvm | 18:22 |
mriedem | doesn't meet the requisite 10 +1s for a +2 | 18:22 |
*** VW_ has joined #openstack-nova | 18:22 | |
JayF | heh | 18:23 |
*** VW__ has quit IRC | 18:23 | |
*** salv-orlando has joined #openstack-nova | 18:23 | |
jroll | mriedem: because we haven't bugged you enough, the backport :P https://review.openstack.org/#/c/176396/ | 18:23 |
*** vilobhmm1 has quit IRC | 18:24 | |
mriedem | jroll: still -1 on that | 18:24 |
mriedem | can you cherry-pick -x that? | 18:24 |
*** vilobhmm1 has joined #openstack-nova | 18:24 | |
*** htruta has joined #openstack-nova | 18:24 | |
mriedem | to preserve the commit hash in the message | 18:24 |
jroll | mriedem: sure thing | 18:24 |
*** dsanders has joined #openstack-nova | 18:24 | |
openstackgerrit | Marian Horban proposed openstack/nova: WIP: Graceful shutdown WSGI server https://review.openstack.org/129167 | 18:25 |
*** penick has joined #openstack-nova | 18:25 | |
*** raildo has joined #openstack-nova | 18:26 | |
mriedem | dansmith: sdague: ^ heh read the first sentence in the commit message | 18:26 |
*** salv-orlando has quit IRC | 18:27 | |
dansmith | I dunno | 18:27 |
dansmith | I guess we should do that | 18:27 |
dansmith | closing the master socket first makes sense | 18:27 |
dansmith | but the API nodes generally are supposed to go away fast when you kill them | 18:27 |
*** jogo has quit IRC | 18:27 | |
dansmith | wonder what alaski thinks of that | 18:27 |
dansmith | might want a timeout or something so we don't stall api exits on a blocked rpc thread for 60 seconds | 18:29 |
jroll | mriedem: updated 176396, even if openstackgerrit doesn't think I did | 18:29 |
mriedem | jroll: the bot doesn't post on stable changes | 18:29 |
jroll | ah | 18:29 |
jroll | thanks | 18:29 |
alaski | dansmith: looking... | 18:30 |
*** ildikov has joined #openstack-nova | 18:30 | |
openstackgerrit | Dan Smith proposed openstack/nova: Add migration_type to Migration database model https://review.openstack.org/181109 | 18:30 |
openstackgerrit | Dan Smith proposed openstack/nova: Add migration_type to Migration object https://review.openstack.org/181110 | 18:30 |
alaski | dansmith: seems reasonable | 18:31 |
*** jichen has joined #openstack-nova | 18:31 | |
alaski | API should go away quickly, but hopefully because they're not doing anything long running | 18:31 |
dansmith | alaski: so that can block for at least a minute on an rpc thing | 18:31 |
*** ericksonfgds has joined #openstack-nova | 18:32 | |
dansmith | alaski: well, if you've just bounced your rabbit, or your computes are down, it can | 18:32 |
*** Marga_ has quit IRC | 18:32 | |
alaski | fair point | 18:32 |
alaski | I like having the option though | 18:32 |
alaski | graceful if you can wait, kill -9 if not | 18:32 |
dansmith | or make sure another TERM will take it out | 18:33 |
alaski | yeah, or that | 18:33 |
dansmith | the problem with closing the socket and then waiting, | 18:33 |
dansmith | is that you end up refusing connections for minutes potentially | 18:34 |
dansmith | if you're managing things with a load balancer, then maybe that's not as much of a problem | 18:34 |
alaski | yeah | 18:34 |
dansmith | but if you're not, this at least requires a heads-up to anyone used to the usual behavior | 18:34 |
alaski | really what we want is a way to bleed connections | 18:34 |
*** eharney_ has joined #openstack-nova | 18:35 | |
alaski | which sort of requires a load balancer | 18:35 |
dansmith | right, was trying to figure out how we'd do that otherwise :) | 18:35 |
*** Marga_ has joined #openstack-nova | 18:36 | |
alaski | so I'm for the feature, but there's a compatibility concern right? | 18:37 |
dansmith | actually | 18:37 |
sdague | mriedem: yeh, I -1ed that patch 2 months ago for the oslo violation, good to know the author cared on that :( | 18:37 |
dansmith | if you drop the master socket, then you could re-start new api workers immediately and the old ones would die off when they're done | 18:37 |
mriedem | sdague: well, wip, so i'm guessing he's using it for testing right now | 18:37 |
sdague | mriedem: maybe | 18:37 |
alaski | dansmith: it seems like the right option would be to have graceful shutdown take another TERM | 18:37 |
mriedem | roman had also pointed out the o-i issue | 18:38 |
*** eharney has quit IRC | 18:38 | |
alaski | sorry, I mean another signal | 18:38 |
dansmith | alaski: what do you mean? | 18:38 |
alaski | to maintain compatibility, keep current behavior and add a way to do the graceful thing | 18:38 |
alaski | it's a bit odd though I suppose | 18:39 |
dansmith | ah | 18:39 |
dansmith | yeah, term is the right signal though | 18:39 |
dansmith | so with the socket closure, | 18:39 |
dansmith | I think the behavior change doesn't matter as much, | 18:39 |
dansmith | because TERM; sleep 1; start-nova-api should still work I think | 18:39 |
dansmith | I hadn't thought it through | 18:39 |
alaski | yeah, if that works then this is great | 18:40 |
dansmith | yeah | 18:40 |
mriedem | i feel really dumb but i think i'm staring at what's failing here and can't figure it out http://logs.openstack.org/66/168966/4/check/gate-nova-tox-functional/8166a8a/console.html#_2015-05-07_17_41_46_380 | 18:40 |
openstackgerrit | Brent Eagles proposed openstack/nova-specs: WIP: VIF plug script support for Nova https://review.openstack.org/162468 | 18:40 |
*** penick has quit IRC | 18:41 | |
*** penick has joined #openstack-nova | 18:41 | |
*** tellesnobrega has quit IRC | 18:41 | |
*** raildo has quit IRC | 18:41 | |
*** ericksonfgds has quit IRC | 18:42 | |
*** htruta has quit IRC | 18:42 | |
*** jogo has joined #openstack-nova | 18:43 | |
*** sdake_ has joined #openstack-nova | 18:43 | |
*** eharney_ is now known as eharney | 18:44 | |
*** exploreshaifali_ has joined #openstack-nova | 18:44 | |
*** exploreshaifali has quit IRC | 18:46 | |
*** sdake has quit IRC | 18:47 | |
*** markmcclain has quit IRC | 18:48 | |
*** jistr has quit IRC | 18:51 | |
*** emagana has quit IRC | 18:53 | |
openstackgerrit | Sudipta Biswas proposed openstack/nova-specs: Report host memory bandwidth as a metric in Nova https://review.openstack.org/180983 | 18:54 |
*** emagana has joined #openstack-nova | 18:55 | |
*** emagana has quit IRC | 18:55 | |
*** tojuvone has quit IRC | 18:56 | |
*** packet has quit IRC | 18:56 | |
kashyap | dansmith: Hi, when you get a moment, w/ 1 week old Nova git, does this look accurate to you: | 18:57 |
*** packet has joined #openstack-nova | 18:57 | |
kashyap | Deleting an instance (from bootable volume) with 'shutdown=REMOVE' -- http://paste.openstack.org/show/216381/ | 18:57 |
dansmith | kashyap: no, doesn't match my experience | 18:58 |
dansmith | kashyap: but I used shutdown=remove not REMOVE | 18:58 |
dansmith | not sure if it matters | 18:58 |
*** Sukhdev has joined #openstack-nova | 18:58 | |
kashyap | Worth a shot, let me retest. | 18:58 |
*** emagana has joined #openstack-nova | 18:59 | |
mriedem | bdm_dict['delete_on_termination'] = (action == 'remove') | 19:00 |
mriedem | kashyap: ^ | 19:00 |
kashyap | mriedem: Darn, I should have looked at the source to begin with :-) | 19:01 |
dansmith | I was looking at docs | 19:01 |
dansmith | where REMOVE is the metavar for "remove or preserve" | 19:01 |
mriedem | yeah docs also say 'remove' | 19:01 |
kashyap | dansmith: Strange, with 'remove' (lower case), after deleting the Nova instance, `cinder list` just hangs?! | 19:02 |
mriedem | http://docs.openstack.org/cli-reference/content/novaclient_commands.html#novaclient_subcommand_boot | 19:02 |
kashyap | No, it doesn't hang, it's just slow. Impatient me. | 19:02 |
kashyap | dansmith: mriedem: Yep, for completness' sake, you're right. Lower case did it (volume was removed, as expected). | 19:03 |
openstackgerrit | Matt Riedemann proposed openstack/nova: Add DB2 support https://review.openstack.org/69047 | 19:03 |
*** VW_ has quit IRC | 19:04 | |
*** VW_ has joined #openstack-nova | 19:07 | |
*** VW_ has quit IRC | 19:07 | |
mriedem | kashyap: i have a mailing address for personal checks | 19:07 |
mriedem | no cash please, i don't trust the mailman | 19:08 |
kashyap | mriedem: :-) | 19:08 |
*** VW_ has joined #openstack-nova | 19:08 | |
openstackgerrit | Jonathan Halterman proposed openstack/nova-specs: Nova VM placement https://review.openstack.org/181117 | 19:09 |
*** markmcclain1 has joined #openstack-nova | 19:09 | |
kashyap | mriedem: BTW, on a unrelated note regarding this patch -- https://review.openstack.org/#/c/171098/, danpb posted his long work (on qemu-devel list) on adding encryption/TLS support for all QEMU I/O channels. | 19:09 |
kashyap | It's being reviewed (albeit a bit slowly) upstream, and /me has been slowly testing it from git, and found a couple of bugs. So I'm sure there's few more iterations in-progress. | 19:11 |
kashyap | (s/in-progress/upcoming) | 19:11 |
*** __gryf has joined #openstack-nova | 19:12 | |
openstackgerrit | Brent Eagles proposed openstack/nova-specs: Distribute PCI Requests Across Multiple Devices https://review.openstack.org/142094 | 19:12 |
kashyap | In short, that avoids the need to tunnel anything via libvirt's RPC layer. | 19:13 |
kashyap | Anyhow, see ya folks for the night! | 19:13 |
mriedem | figured out my functional test failure from 30 minutes ago | 19:14 |
mriedem | thanks to no one's help god! | 19:14 |
openstackgerrit | Matt Riedemann proposed openstack/nova: Show 'reserved' status in os-fixed-ips https://review.openstack.org/168966 | 19:14 |
*** zul has quit IRC | 19:16 | |
*** markmcclain1 is now known as markmcclain | 19:16 | |
melwitt | mriedem: I looked but didn't have anything helpful to say. :P that response would have gotten to that point as not a string. what was it? | 19:17 |
mriedem | melwitt: i'm guessing an error or something dumb? bool? | 19:17 |
dansmith | is melwitt core yet? | 19:18 |
*** sudipto has quit IRC | 19:18 | |
mriedem | from the number of random +1s in the ML, yes i think so | 19:19 |
dansmith | well, I meant has it become official | 19:19 |
dansmith | i.e. can she +2 my patches yet | 19:19 |
melwitt | mriedem: I got confused by the series of json.loads and json.dumps before that so I was more puzzled how it got through that as not a string | 19:19 |
dansmith | it's clear the +1 store is out of stock | 19:19 |
*** tellesnobrega has joined #openstack-nova | 19:20 | |
melwitt | lol :P | 19:20 |
mriedem | i think there is a 1 week policy on people speaking now or forever holding their peace | 19:21 |
dansmith | right, | 19:21 |
dansmith | and that was up today | 19:21 |
mriedem | oh | 19:21 |
dansmith | but I think johnthetubaguy should do the add | 19:21 |
mriedem | well | 19:21 |
mriedem | he's going to be late | 19:21 |
dansmith | heh | 19:22 |
melwitt | I see "AssertionError: False is not true : Timed out waiting for 10.100.0.4 to become reachable" appeared in the gate recently (5/4) only 55 hits in the past 7 days | 19:22 |
openstackgerrit | Jonathan Halterman proposed openstack/nova-specs: Nova VM placement https://review.openstack.org/181117 | 19:23 |
openstackgerrit | Brent Eagles proposed openstack/nova-specs: Refactor allocate_for_instance https://review.openstack.org/141129 | 19:23 |
mriedem | melwitt: link? | 19:23 |
melwitt | mriedem: http://logstash.openstack.org/#eyJzZWFyY2giOiJtZXNzYWdlOlwiRmFsc2UgaXMgbm90IHRydWUgOiBUaW1lZCBvdXQgd2FpdGluZyBmb3JcIiBBTkQgdGFnczpcImNvbnNvbGVcIiIsImZpZWxkcyI6WyJidWlsZF9zaG9ydF91dWlkIiwiYnVpbGRfc3RhdHVzIiwiYnVpbGRfcXVldWUiLCJmaWxlbmFtZSJdLCJvZmZzZXQiOjAsInRpbWVmcmFtZSI6IjYwNDgwMCIsImdyYXBobW9kZSI6ImNvdW50IiwidGltZSI6eyJ1c2VyX2ludGVydmFsIjowfSwic3RhbXAiOjE0MzEwMjYxNTY2MTB9 | 19:24 |
mriedem | melwitt: well, | 19:25 |
*** angdraug has quit IRC | 19:25 | |
mriedem | some garbage queues in there | 19:25 |
mriedem | restrict to gate and it's 1 hit | 19:25 |
mriedem | on a barbicanclient job | 19:26 |
mriedem | so, you know | 19:26 |
mriedem | quick, how do i tell a guy i don't want him to come to my office to talk about a thing i don't want to talk about | 19:26 |
*** raildo has joined #openstack-nova | 19:27 | |
*** Marga_ has quit IRC | 19:28 | |
openstackgerrit | Brent Eagles proposed openstack/nova-specs: Refactor of the Neutron network adapter https://review.openstack.org/131413 | 19:29 |
dansmith | mriedem: start farting | 19:30 |
tjones1 | mriedem: tell him you have a sore throat and may be contagious | 19:30 |
dansmith | tjones1: oh sure, come up with a PG solution | 19:30 |
mriedem | dansmith: gas was my inital thought as well :) | 19:30 |
dansmith | of course | 19:30 |
tjones1 | dansmith: boys boys…. | 19:30 |
melwitt | mriedem: yeah, I noticed it was only one gate hit. I saw it pop up on a change I've been reviewing, one that should be fine. should I just recheck it? or open a bug, no e-r query? | 19:30 |
tjones1 | gas is always your initial thought | 19:30 |
dansmith | tjones1: hey, don't ... yeah, nevermind | 19:30 |
tjones1 | lol | 19:31 |
mriedem | 'you'll have to speak up, i'm wearing a towel' | 19:31 |
*** zul has joined #openstack-nova | 19:31 | |
mriedem | https://www.youtube.com/watch?v=jk8SToEQPGw | 19:31 |
mriedem | melwitt: let me check logstash again | 19:32 |
mriedem | on jobs we don't expect to suck | 19:32 |
openstackgerrit | Dan Smith proposed openstack/nova: Refactor tests that use compute's deprecated run_instance() method https://review.openstack.org/179950 | 19:32 |
openstackgerrit | Dan Smith proposed openstack/nova: Drop compute RPC 3.x support https://review.openstack.org/179951 | 19:32 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SpawnFixture https://review.openstack.org/179949 | 19:32 |
openstackgerrit | Dan Smith proposed openstack/nova: Switch to _set_instance_obj_error_state in build_and_run_instance https://review.openstack.org/180291 | 19:32 |
openstackgerrit | Dan Smith proposed openstack/nova: Update kilo version alias https://review.openstack.org/181127 | 19:32 |
*** bkopilov has quit IRC | 19:33 | |
*** busterswt has quit IRC | 19:34 | |
__gryf | dansmith, hi | 19:36 |
dansmith | __gryf: I still haven't gone back to make a full pass | 19:37 |
__gryf | dansmith, i just have a qst regrding api | 19:37 |
dansmith | alright :) | 19:37 |
__gryf | dansmith, so if the v2 is frozen and we have to use v2.1 how;s that map to extensions v2? | 19:38 |
*** ociuhandu has quit IRC | 19:38 | |
__gryf | i mean | 19:38 |
dansmith | parse error. | 19:39 |
__gryf | where to put the actual implementation | 19:39 |
dansmith | in the v2.1 code | 19:39 |
__gryf | since with v2 extension (which os-services are) it was pretty obvious | 19:39 |
dansmith | cleverly hiding in the v3/ directory in the tree | 19:39 |
dansmith | __gryf: extensions are dead | 19:39 |
__gryf | cleverly… | 19:39 |
__gryf | that's the last word… oh, wait | 19:40 |
__gryf | so, shoul I move all the needed things out of api/openstack/compute/contrib/services.py to api/openstack/compute/schemas/v3/services.py ? | 19:41 |
dansmith | yes | 19:41 |
__gryf | ack | 19:41 |
dansmith | well, | 19:41 |
dansmith | wait, schemas | 19:41 |
__gryf | yeah, my bad | 19:42 |
dansmith | I think you mean api/openstack/compute/plugins/v3/services.py | 19:42 |
__gryf | right | 19:42 |
dansmith | then: yes :) | 19:42 |
__gryf | and what about all the other api calls? | 19:42 |
__gryf | enable/disable/details and so on? | 19:42 |
dansmith | all what other calls? | 19:42 |
openstackgerrit | Jonathan Halterman proposed openstack/nova-specs: Nova VM placement https://review.openstack.org/181117 | 19:43 |
*** bkopilov has joined #openstack-nova | 19:43 | |
__gryf | api calls, which currently resides in v2 ext? | 19:43 |
*** ildikov has quit IRC | 19:43 | |
__gryf | for os-services | 19:43 |
__gryf | should also be moved within that BP? | 19:43 |
dansmith | nothing existing should be moved, | 19:44 |
dansmith | you should just make your additions to the v3 one | 19:44 |
__gryf | ok. | 19:44 |
__gryf | dansmith, doesn't it feel a bit wierd if the call for certain thing, which is a part of the group (like the os-services) will be completely different, than the other ones? | 19:49 |
dansmith | __gryf: not sure what you mean | 19:50 |
__gryf | ok. let me explain | 19:50 |
*** yamahata has quit IRC | 19:50 | |
*** gokrokve has joined #openstack-nova | 19:51 | |
__gryf | now we have several calls to interact with services, where the url looks like this /v2/{tenant}/os-services[/action] and with the defined payload, which is similar for all the update actions | 19:53 |
*** bkopilov has quit IRC | 19:53 | |
*** bkopilov has joined #openstack-nova | 19:53 | |
jaypipes | melwitt: https://review.openstack.org/#/c/159626/ could use your new +2 :) | 19:54 |
__gryf | and yet if I introduce another call, which might looks like this /v2.1/os-services/{service-id} with appropriate payload with all needed data | 19:54 |
__gryf | it might confuse user | 19:55 |
*** heyongli has quit IRC | 19:55 | |
*** gokrokve has quit IRC | 19:55 | |
*** marun has quit IRC | 19:56 | |
*** b3nt_pin has quit IRC | 19:56 | |
__gryf | dansmith, my point is, that there will be a bug difference between these calls | 19:56 |
__gryf | s/bug/big/ | 19:56 |
dansmith | __gryf: for v2 and v2.1 you mean? | 19:56 |
*** busterswt has joined #openstack-nova | 19:56 | |
__gryf | yes | 19:56 |
dansmith | __gryf: v2 is going away real soon | 19:56 |
__gryf | within the same group | 19:56 |
melwitt | jaypipes: cool, I will revisit after I'm added :) | 19:56 |
dansmith | __gryf: so, that's fine | 19:56 |
*** angdraug has joined #openstack-nova | 19:57 | |
__gryf | dansmith, if v2 is going away, than i assume, that all functionalty in v2 have to be translated to v2.1, am I rihgt? | 19:58 |
dansmith | __gryf: *has been. yes | 19:58 |
*** bkopilov has quit IRC | 19:59 | |
*** ildikov has joined #openstack-nova | 19:59 | |
*** bkopilov has joined #openstack-nova | 19:59 | |
__gryf | dansmith, hm. if so, than I should not see any /v2/…/os-services while using --debug i nova cmdline util, right? | 20:00 |
*** VW_ has quit IRC | 20:00 | |
__gryf | or it's a nova-client fault for still using v2? | 20:00 |
dansmith | __gryf: are you like trying to prove me wrong, or are you asking for actual guidance? | 20:01 |
*** matrohon has joined #openstack-nova | 20:01 | |
jaypipes | melwitt: :( oh, sorry, I thought you were already added? | 20:01 |
dansmith | melwitt: I assume that novaclient is fully v2.1ified because v2.1 was released as the current version in kilo, is that right? | 20:01 |
*** heyongli has joined #openstack-nova | 20:01 | |
__gryf | dansmith, I'm just trying to understand how it works | 20:01 |
*** salv-orlando has joined #openstack-nova | 20:01 | |
dansmith | _gryf: http://developer.openstack.org/api-ref.html | 20:02 |
dansmith | __gryf: see how it says v2.1 is "current" | 20:02 |
*** annegentle has quit IRC | 20:02 | |
*** annegentle has joined #openstack-nova | 20:02 | |
melwitt | dansmith: it's not actually. there are patches up to do it. let me see | 20:03 |
*** achanda has quit IRC | 20:04 | |
dansmith | it only really matters that we have that before we mark as deprecated, AFAIK | 20:04 |
*** achanda has joined #openstack-nova | 20:04 | |
__gryf | dansmith, yes, I can see that. I also see, that compute API v2.1 lack of os-services, while compute api v2 extensions have it | 20:04 |
melwitt | okay, just checked it's in the same state. alex_xu had some comments on how those patches should be changed a bit. we have had the API experts going through it | 20:04 |
bauzas | nova meeting in 55 mins right? | 20:05 |
edleafe | bauzas: yep | 20:05 |
dansmith | __gryf: I see services in v3 | 20:05 |
bauzas | _gryf: os-services is also in v2.1 AFAICS | 20:05 |
*** heyongli has quit IRC | 20:05 | |
bauzas | _gryf: but the docs is not good | 20:05 |
dansmith | __gryf: https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/plugins/v3/services.py | 20:06 |
bauzas | _gryf: that's something I already discussed with alex_xu | 20:06 |
__gryf | bauzas, oh. I see. bummer | 20:06 |
*** VW_ has joined #openstack-nova | 20:06 | |
__gryf | dansmith, thx for pointing it out (: | 20:06 |
bauzas | _gryf: your point is valid, the docs need updating | 20:06 |
__gryf | bauzas, so it seems like bug… in documentation | 20:07 |
bauzas | _gryf: not really | 20:07 |
*** tjones2 has joined #openstack-nova | 20:07 | |
bauzas | _gryf: that's in Nova tree | 20:07 |
dansmith | yeah | 20:07 |
*** baoli has joined #openstack-nova | 20:07 | |
__gryf | dansmith, bauzas, thanks for clarifying (: | 20:07 |
melwitt | __gryf: so by default, novaclient will go to the keystone catalog endpoint named "compute". if you want it to call elsewhere, you can tell it by --service-type <name> where you have v2.1 being served. it might probably work | 20:07 |
*** alexpilotti has quit IRC | 20:08 | |
melwitt | if you're using devstack, it will be --service-type "computev21" and then you should see novaclient call to the /v2.1/... | 20:08 |
*** beagles has joined #openstack-nova | 20:08 | |
bauzas | melwitt: what's the status of it for v2.1 ? is it still v2 the default for the client ? | 20:08 |
__gryf | melwitt, ok, I'll try it out, thanks | 20:08 |
*** tjones1 has quit IRC | 20:08 | |
*** otter768 has joined #openstack-nova | 20:08 | |
bauzas | melwitt: or is it v2.1 but not micro-versioned ? | 20:08 |
*** achanda has quit IRC | 20:08 | |
*** baoli has quit IRC | 20:08 | |
melwitt | bauzas: so, the default is whatever is being served out of "compute" in the service catalog. so it would be deployment dependent. if it's devstack, "compute" == v2 | 20:09 |
dansmith | bauzas: right, v2.1 should work for v2 clients, that's the point | 20:09 |
*** alexpilotti has joined #openstack-nova | 20:09 | |
*** baoli has joined #openstack-nova | 20:09 | |
dansmith | so it should be fine to point the client at it, AFAIK | 20:09 |
melwitt | bauzas: yeah it's just it can't do anything with micro-versions yet | 20:09 |
bauzas | dansmith: I've been heard that input validation is stricter for v2.1 | 20:10 |
dansmith | bauzas: sure | 20:10 |
*** heyongli has joined #openstack-nova | 20:10 | |
bauzas | melwitt: dansmith: okay I see | 20:10 |
*** toure_ has joined #openstack-nova | 20:10 | |
*** toure_ has quit IRC | 20:11 | |
* melwitt runs to lunch | 20:11 | |
bauzas | melwitt: while you're in, do you have a few mins to discuss on https://review.openstack.org/#/c/149562/ ? | 20:11 |
bauzas | ergh :) | 20:11 |
bauzas | melwitt: ttyl then :) | 20:11 |
*** baoli has quit IRC | 20:13 | |
*** otter768 has quit IRC | 20:13 | |
*** baoli has joined #openstack-nova | 20:13 | |
*** alexpilotti has quit IRC | 20:16 | |
*** salv-orlando has quit IRC | 20:16 | |
*** Longgeek has quit IRC | 20:18 | |
*** alexpilotti has joined #openstack-nova | 20:21 | |
*** Mike_D_laptop has quit IRC | 20:21 | |
*** VW_ has quit IRC | 20:21 | |
*** Marga_ has joined #openstack-nova | 20:23 | |
*** Marga_ has quit IRC | 20:23 | |
*** Marga_ has joined #openstack-nova | 20:24 | |
*** exploreshaifali_ has quit IRC | 20:26 | |
*** gokrokve has joined #openstack-nova | 20:26 | |
*** oro has quit IRC | 20:27 | |
*** gokrokve has quit IRC | 20:28 | |
*** gokrokve has joined #openstack-nova | 20:28 | |
*** alexpilotti has quit IRC | 20:29 | |
*** dsanders has quit IRC | 20:29 | |
*** nikhil_k-afk is now known as nikhil_k | 20:29 | |
*** sdake_ has quit IRC | 20:29 | |
*** baoli has quit IRC | 20:29 | |
*** emagana has quit IRC | 20:31 | |
*** __gryf has quit IRC | 20:31 | |
*** emagana has joined #openstack-nova | 20:32 | |
*** tonytan4ever has quit IRC | 20:33 | |
*** GLaupre has joined #openstack-nova | 20:33 | |
*** zzzeek has quit IRC | 20:35 | |
*** baoli has joined #openstack-nova | 20:35 | |
*** zzzeek has joined #openstack-nova | 20:35 | |
*** ericksonfgds has joined #openstack-nova | 20:35 | |
*** bnemec has quit IRC | 20:36 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Run using rootwrap daemon as True https://review.openstack.org/180809 | 20:38 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add rootwrap daemon mode support (Switched off) https://review.openstack.org/180695 | 20:38 |
*** jichen has quit IRC | 20:41 | |
*** Marga_ has quit IRC | 20:41 | |
*** Marga_ has joined #openstack-nova | 20:42 | |
*** Nikolay_St has joined #openstack-nova | 20:42 | |
*** dsanders has joined #openstack-nova | 20:43 | |
*** achanda has joined #openstack-nova | 20:44 | |
*** vladikr has quit IRC | 20:44 | |
*** marun has joined #openstack-nova | 20:44 | |
openstackgerrit | Dan Smith proposed openstack/nova: Refactor tests that use compute's deprecated run_instance() method https://review.openstack.org/179950 | 20:48 |
openstackgerrit | Dan Smith proposed openstack/nova: Drop compute RPC 3.x support https://review.openstack.org/179951 | 20:48 |
openstackgerrit | Dan Smith proposed openstack/nova: Update kilo version alias https://review.openstack.org/181127 | 20:48 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SpawnFixture https://review.openstack.org/179949 | 20:48 |
openstackgerrit | Dan Smith proposed openstack/nova: Switch to _set_instance_obj_error_state in build_and_run_instance https://review.openstack.org/180291 | 20:48 |
*** VW_ has joined #openstack-nova | 20:51 | |
*** bnemec has joined #openstack-nova | 20:51 | |
*** VW_ has quit IRC | 20:53 | |
*** VW__ has joined #openstack-nova | 20:53 | |
mriedem | nova meeting in 6 minutes | 20:54 |
mriedem | get ready to argue about mid cycle meetup schedule | 20:54 |
*** andrewbogott has left #openstack-nova | 20:54 | |
dansmith | no way | 20:55 |
dansmith | that's done | 20:55 |
dansmith | no arguing | 20:55 |
*** baoli has quit IRC | 20:55 | |
*** baoli has joined #openstack-nova | 20:56 | |
*** oro has joined #openstack-nova | 20:57 | |
*** moshele has quit IRC | 20:58 | |
*** VW__ has quit IRC | 20:58 | |
*** Marga_ has quit IRC | 20:58 | |
*** Sukhdev has quit IRC | 20:58 | |
*** browne has quit IRC | 20:58 | |
*** Marga_ has joined #openstack-nova | 20:58 | |
bauzas | mriedem: I can invite in my house, fine ? | 20:59 |
bauzas | mriedem: for the midcycle | 20:59 |
*** VW_ has joined #openstack-nova | 21:01 | |
*** thangp has quit IRC | 21:04 | |
*** annegentle has quit IRC | 21:04 | |
*** aysyd has quit IRC | 21:04 | |
*** exploreshaifali_ has joined #openstack-nova | 21:06 | |
*** sdake has joined #openstack-nova | 21:08 | |
*** jog0 has joined #openstack-nova | 21:09 | |
*** jog0 has quit IRC | 21:09 | |
*** jog0 has joined #openstack-nova | 21:09 | |
*** jogo has quit IRC | 21:11 | |
*** sdake_ has joined #openstack-nova | 21:11 | |
*** Anticime1 is now known as Anticimex | 21:14 | |
*** sdake has quit IRC | 21:15 | |
*** EmilienM is now known as EmilienM|afk | 21:16 | |
*** baoli has quit IRC | 21:16 | |
*** baoli has joined #openstack-nova | 21:17 | |
*** dprince has quit IRC | 21:20 | |
thomasem | Anyone got a moment to chat about novaclient? | 21:22 |
*** packet has quit IRC | 21:23 | |
*** rms_13 has joined #openstack-nova | 21:23 | |
*** rms_13 has left #openstack-nova | 21:23 | |
*** claudiub has joined #openstack-nova | 21:24 | |
*** rms_13 has joined #openstack-nova | 21:24 | |
*** neelashah has joined #openstack-nova | 21:25 | |
*** bnemec has quit IRC | 21:25 | |
*** tjones2 has quit IRC | 21:29 | |
*** sdake_ is now known as sdake | 21:30 | |
*** ijw has joined #openstack-nova | 21:32 | |
*** Sukhdev has joined #openstack-nova | 21:32 | |
*** redbeard has quit IRC | 21:32 | |
openstackgerrit | Dan Smith proposed openstack/nova: Drop compute RPC 3.x support https://review.openstack.org/179951 | 21:33 |
*** IanGovett has quit IRC | 21:33 | |
mikal | Can I be Sarcasm Czar? | 21:35 |
*** tjones1 has joined #openstack-nova | 21:35 | |
*** exploreshaifali_ has quit IRC | 21:35 | |
*** MVenesio has quit IRC | 21:37 | |
*** baoli has quit IRC | 21:38 | |
*** baoli has joined #openstack-nova | 21:39 | |
*** jog0 is now known as jogo | 21:41 | |
*** jogo has quit IRC | 21:41 | |
*** jogo has joined #openstack-nova | 21:43 | |
*** rfolco has quit IRC | 21:45 | |
*** mwagner_lap has quit IRC | 21:45 | |
*** matrohon has quit IRC | 21:47 | |
*** baoli has quit IRC | 21:49 | |
*** baoli has joined #openstack-nova | 21:50 | |
*** Marga_ has quit IRC | 21:50 | |
jaypipes | thomasem: we're all in the Nova IRC meeting right this moment :) but leave your question here!@ | 21:50 |
*** isd has quit IRC | 21:51 | |
dims | mikal: johnthetubaguy: please restore https://review.openstack.org/#/c/105404/ i'd like to pick that up for liberty | 21:52 |
*** oro has quit IRC | 21:53 | |
mikal | dims: I don't think we have to? You just upload a new revision of that change id and it opens up magically | 21:53 |
dims | not if it is abandoned mikal | 21:54 |
dims | ! [remote rejected] HEAD -> refs/publish/master/bp/rootwrap-daemon-mode (change 105404 closed) | 21:54 |
dims | error: failed to push some refs to 'ssh://dims-v@review.openstack.org:29418/openstack/nova-specs.git' | 21:54 |
mikal | Huh | 21:54 |
*** sdake_ has joined #openstack-nova | 21:54 | |
dims | at least for non cores it does not work | 21:54 |
mikal | I have misunderstoof abandon then | 21:55 |
sc68cal | OK, so I added a new section to the cross project page - hopefully my table makes senes to everyone https://wiki.openstack.org/wiki/CrossProjectLiaisons#Inter-project_Liaisons | 21:55 |
mikal | Which I think means my email to -dev the other day was a pack of lies | 21:55 |
lifeless | you need to unabandon | 21:55 |
lifeless | then you can push | 21:55 |
dims | or probably because this was not mine | 21:55 |
lifeless | and infra, -core and the owner can unabandon | 21:55 |
dims | to begin with | 21:55 |
mikal | dims: done | 21:55 |
dims | thanks much mikal | 21:56 |
mikal | NP | 21:56 |
mikal | Ahhh, ok | 21:56 |
mikal | Its because you're not the owner | 21:56 |
mikal | That makes my email a smaller pack of lies | 21:56 |
mikal | Which is good | 21:56 |
lifeless | I've had that error when I'm the owner | 21:56 |
*** david-lyle has joined #openstack-nova | 21:56 | |
lifeless | I think we probably need to test; I'm not sure if that bit is acl controlled or intrinsic | 21:56 |
*** burt has quit IRC | 21:57 | |
*** neelashah has quit IRC | 21:58 | |
*** sdake has quit IRC | 21:58 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova-specs: Add rootwrap-daemon-mode blueprint https://review.openstack.org/105404 | 22:00 |
mikal | dansmith: tonyb means insufferable in the "rides his bike to the gym and then tweets about it" sense | 22:00 |
*** gokrokve has quit IRC | 22:01 | |
*** gokrokve has joined #openstack-nova | 22:01 | |
*** andreykurilin__ has joined #openstack-nova | 22:01 | |
tonyb | mikal: and the "my body is a temple" sense | 22:01 |
dansmith | mikal: oh | 22:01 |
mikal | tonyb: true dat | 22:01 |
mikal | tonyb: a fatty fatty temple | 22:01 |
dansmith | my body is the size of a temple, does that work? | 22:01 |
dansmith | heh | 22:01 |
mikal | tonyb: I miss my bikes, please go for a ride for me | 22:02 |
* mikal is stuck in Sydney as a punishment for his sins | 22:02 | |
tonyb | mikal: I went last night | 22:02 |
*** artom has quit IRC | 22:04 | |
*** RuiChen has quit IRC | 22:07 | |
*** gokrokve has quit IRC | 22:07 | |
*** emagana has quit IRC | 22:07 | |
*** RuiChen has joined #openstack-nova | 22:07 | |
*** eharney has quit IRC | 22:07 | |
*** gokrokve has joined #openstack-nova | 22:07 | |
*** Marga_ has joined #openstack-nova | 22:08 | |
*** otter768 has joined #openstack-nova | 22:09 | |
*** Nic has quit IRC | 22:10 | |
edleafe | mikal: I have a 103-mile ride on Saturday. I'll think of you the whole way | 22:10 |
dansmith | edleafe: creepy. nice going. | 22:10 |
*** baoli has quit IRC | 22:10 | |
edleafe | dansmith: I try | 22:10 |
*** baoli has joined #openstack-nova | 22:11 | |
*** nelsnelson has quit IRC | 22:11 | |
*** claudiub has quit IRC | 22:12 | |
*** dboik_ has joined #openstack-nova | 22:12 | |
*** otter768 has quit IRC | 22:14 | |
mikal | edleafe: you are dead to me because you're better at riding than me | 22:15 |
edleafe | mikal: if I survive it, that is | 22:15 |
*** browne has joined #openstack-nova | 22:16 | |
mikal | Oh, good point | 22:16 |
*** dboik has quit IRC | 22:16 | |
mikal | If you fail massive after say three miles, then you're definitely alive to me | 22:16 |
*** dboik_ has quit IRC | 22:17 | |
edleafe | mikal: Hmmm... I certainly plan on making it through at least five... | 22:19 |
*** ddaskal has quit IRC | 22:19 | |
*** tjones1 has quit IRC | 22:20 | |
*** gokrokve has quit IRC | 22:21 | |
*** baoli has quit IRC | 22:21 | |
mriedem | dims: when the functional tests run, we get this deprecation warning: http://paste.openstack.org/show/216526/ | 22:21 |
mriedem | dims: i'm trying to find out what's using oslo.middleware, nova/etc/api-paste.ini was the only thing i saw | 22:22 |
*** baoli has joined #openstack-nova | 22:22 | |
mriedem | i tried changing that to oslo_middleware in there but that doesn't fix anythin | 22:22 |
mikal | edleafe: that's border line then | 22:22 |
*** sdake has joined #openstack-nova | 22:22 | |
*** salv-orlando has joined #openstack-nova | 22:25 | |
*** Marga_ has quit IRC | 22:25 | |
*** sdake_ has quit IRC | 22:25 | |
*** Marga_ has joined #openstack-nova | 22:26 | |
*** kylek3h has quit IRC | 22:26 | |
*** Marga_ has quit IRC | 22:29 | |
*** Marga_ has joined #openstack-nova | 22:29 | |
*** tjones1 has joined #openstack-nova | 22:30 | |
*** salv-orlando has quit IRC | 22:31 | |
*** sambol has quit IRC | 22:32 | |
*** ociuhandu has joined #openstack-nova | 22:34 | |
*** changbl has quit IRC | 22:36 | |
*** Raildo_ has joined #openstack-nova | 22:36 | |
*** david-lyle has quit IRC | 22:38 | |
*** IanGovett has joined #openstack-nova | 22:39 | |
*** promulo__ has quit IRC | 22:46 | |
*** ijw has quit IRC | 22:46 | |
*** promulo_ has joined #openstack-nova | 22:46 | |
*** promulo_ has quit IRC | 22:46 | |
*** promulo_ has joined #openstack-nova | 22:46 | |
mriedem | oh barf | 22:48 |
mriedem | copy_qcow2_image takes 76 seconds on first boot | 22:48 |
mriedem | rpc_response_timeout is 60 seconds | 22:48 |
*** ujjain has quit IRC | 22:48 | |
mriedem | barf | 22:48 |
*** sdake has quit IRC | 22:51 | |
*** ujjain has joined #openstack-nova | 22:52 | |
*** ujjain has quit IRC | 22:52 | |
*** ujjain has joined #openstack-nova | 22:52 | |
*** sdake has joined #openstack-nova | 22:52 | |
*** baoli has quit IRC | 22:53 | |
*** baoli has joined #openstack-nova | 22:54 | |
*** BadCub is now known as BadCub_Hiding | 22:55 | |
mriedem | does anyone use preallocate_images=space? | 23:03 |
mriedem | config help on that one isn't very helpful | 23:04 |
*** baoli has quit IRC | 23:04 | |
GLaupre | Hello, I have this blueprint https://blueprints.launchpad.net/nova/+spec/improve-vnic-sysfs-parsing, I am working on the code and on the spec. That do I need to make the blueprint accepted? Milestone? | 23:07 |
*** zsais22 has joined #openstack-nova | 23:10 | |
*** zsais222 has quit IRC | 23:13 | |
mriedem | GLaupre: you have to have the spec up | 23:14 |
mriedem | and approved first | 23:14 |
mriedem | before you set milestones, but i don't think we're setting milestones this release b/c it's pointless | 23:14 |
*** zsais22 has quit IRC | 23:16 | |
*** VW_ has quit IRC | 23:17 | |
*** hemna is now known as hemnafk | 23:17 | |
*** dboik has joined #openstack-nova | 23:23 | |
*** dboik_ has joined #openstack-nova | 23:25 | |
*** markvoelker has quit IRC | 23:26 | |
*** dboik has quit IRC | 23:28 | |
*** Marga_ has quit IRC | 23:28 | |
*** park_heijlong has joined #openstack-nova | 23:31 | |
*** Sukhdev has quit IRC | 23:32 | |
park_heijlong | ping melwitt | 23:32 |
melwitt | park_heijlong: hi | 23:33 |
park_heijlong | melwitt, I sent a email to your yahoo-inc, need sync with you about the bug1420192 | 23:35 |
park_heijlong | could you help to take a view...thanks | 23:35 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/181212 | 23:36 |
melwitt | park_heijlong: ah, right. it's usually better to comment on the bug it's easier for me to track that way. one moment | 23:36 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/python-novaclient: Updated from global requirements https://review.openstack.org/181238 | 23:37 |
park_heijlong | melwitt, ok, just need sync with you before update the bug. will update the bug next :-) | 23:37 |
melwitt | park_heijlong: okay so how to reproduce the issue. are you using neutron? because I see the stack trace when I use nova-network | 23:38 |
park_heijlong | melwitt, I use nova-network too | 23:39 |
park_heijlong | melwitt, just the n-api log seems different | 23:39 |
melwitt | park_heijlong: hm, yeah. when I do the same command 'nova interface-attach vm1' I get 400 from nova too | 23:41 |
mriedem | ha | 23:42 |
mriedem | remove_unused_kernels | 23:42 |
dansmith | melwitt: interface-attach isn't supported in nova-net is it? | 23:42 |
mriedem | defaults to false, says we should default to true at some point, looks like that was added back in grizzly | 23:42 |
dansmith | there was talk of adding it, but I didn't think that happened | 23:42 |
*** Marga_ has joined #openstack-nova | 23:42 | |
*** busterswt has quit IRC | 23:42 | |
mriedem | i think vishy proposed that but it got held up by paperwork | 23:42 |
park_heijlong | melwitt, agree with dansmith | 23:42 |
melwitt | dansmith: it's not, park_heijlong gets a 501 network driver doesn't support, but I see a 400 and a stack trace in n-api.log | 23:42 |
dansmith | ah | 23:43 |
dims | mriedem: does it happen in any of the gate/check jobs? | 23:43 |
dansmith | melwitt: also, in other obvious news, the sky is blue? | 23:43 |
dansmith | er s/?/!/ | 23:43 |
melwitt | lol | 23:44 |
mriedem | dims: which? | 23:44 |
park_heijlong | melwitt: make sure we are using the same branch.... I use master | 23:44 |
dims | mriedem: the DeprecationWarning for oslo_middleware | 23:45 |
dansmith | park_heijlong: I get a 501 as well | 23:45 |
melwitt | uh oh :( | 23:46 |
dansmith | heh | 23:46 |
park_heijlong | dansmiith: y, :), inerface-attach is NOT support by nova-network at this moment... | 23:46 |
melwitt | so I must sheepishly admit my environment has cells enabled. but I didn't think about it affecting this | 23:46 |
dansmith | oh my | 23:47 |
dims | brb | 23:47 |
dansmith | if so, that's terrible | 23:47 |
* melwitt hides | 23:47 | |
dansmith | not your fault, | 23:48 |
mriedem | dims: yes http://logs.openstack.org/66/168966/5/check/gate-nova-tox-functional/36ca709/console.html#_2015-05-07_19_35_03_749 | 23:48 |
dansmith | it's just really not ideal if we get different behavior on that level between cells and non-cells | 23:48 |
*** dims_ has joined #openstack-nova | 23:50 | |
melwitt | yeah... let me figure out how this is doing different, doesn't really make sense to me | 23:51 |
*** dims has quit IRC | 23:52 | |
*** mriedem is now known as mriedem_away | 23:54 | |
*** takedakn has joined #openstack-nova | 23:55 | |
*** markvoelker has joined #openstack-nova | 23:56 | |
*** markvoelker has quit IRC | 23:57 | |
*** markvoelker has joined #openstack-nova | 23:57 | |
dims_ | dansmith: melwitt: mriedem_away: anyone have some time to give feedback on a devstack review to enable running nova-api and nova-ec2-api under Apache? (has 2 +2 | 23:59 |
dims_ | s from devstack core) | 23:59 |
dansmith | I have no idea if that works | 23:59 |
dims_ | dansmith: i have a companion review that enables the switch and runs all the jobs | 23:59 |
dansmith | are they asking for nova feedback? | 23:59 |
dims_ | y | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!