| *** markvoelker has joined #openstack-nova | 00:00 | |
| *** r-daneel has quit IRC | 00:01 | |
| *** mtanino has joined #openstack-nova | 00:03 | |
| *** xyang1 has quit IRC | 00:06 | |
| *** sorrison_laptop has joined #openstack-nova | 00:06 | |
| *** Longgeek has joined #openstack-nova | 00:13 | |
| *** dave-mccowan has quit IRC | 00:14 | |
| *** asettle has joined #openstack-nova | 00:15 | |
| *** tpeoples has quit IRC | 00:15 | |
| *** hemna is now known as hemnafk | 00:19 | |
| *** claudiub has joined #openstack-nova | 00:19 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/nova-specs: Add multiple compute host support for Ironic driver https://review.openstack.org/194453 | 00:19 |
|---|---|---|
| openstackgerrit | xing-yang proposed openstack/nova: Add os-brick based LibvirtVolumeDriver for ScaleIO https://review.openstack.org/194454 | 00:20 |
| *** lyarwood has quit IRC | 00:22 | |
| *** lyarwood has joined #openstack-nova | 00:22 | |
| *** doude_ has joined #openstack-nova | 00:27 | |
| *** doude has quit IRC | 00:28 | |
| *** jasondotstar has quit IRC | 00:29 | |
| *** jamielennox|away is now known as jamielennox | 00:29 | |
| *** isd has joined #openstack-nova | 00:30 | |
| *** eliqiao1 has left #openstack-nova | 00:35 | |
| *** browne has quit IRC | 00:35 | |
| *** browne has joined #openstack-nova | 00:36 | |
| *** ociuhandu has quit IRC | 00:38 | |
| *** mriedem has quit IRC | 00:43 | |
| *** electrocucaracha has quit IRC | 00:44 | |
| *** nelsnelson has quit IRC | 00:44 | |
| *** isd has quit IRC | 00:45 | |
| *** Longgeek has quit IRC | 00:46 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Add common function for v2.1 API flavor_get. https://review.openstack.org/136568 | 00:47 |
| openstackgerrit | Matt Riedemann proposed openstack/nova: cells: add devstack/tempest-dsvm-cells-rc for gating https://review.openstack.org/194411 | 00:48 |
| *** pixelb has joined #openstack-nova | 00:48 | |
| *** nihilifer has quit IRC | 00:51 | |
| *** vilobhmm has joined #openstack-nova | 00:51 | |
| mikal | I just re-realized that we're using an orphaned import of oslo-incubator's policy module | 00:52 |
| *** yingjun has joined #openstack-nova | 00:52 | |
| mikal | Does anyone remember if someone is working on porting us to the graduated thing? | 00:52 |
| mikal | dims: ^--- | 00:52 |
| *** ijw_ has quit IRC | 00:54 | |
| *** bradjones has quit IRC | 00:54 | |
| *** ijw_ has joined #openstack-nova | 00:54 | |
| *** dims has quit IRC | 00:55 | |
| claudiub | mikal: hi. I was wondering if you have some time to take a look at the hyper-v cluster spec? :) https://review.openstack.org/#/c/188374/ | 00:56 |
| *** dims has joined #openstack-nova | 00:57 | |
| mikal | claudiub: I am pretty sure I already have that open in a tab. I'll try and take a look sometime later today. | 00:57 |
| claudiub | mikal: awesome! thanks! :D | 00:58 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Not pass secgroup values if not specified https://review.openstack.org/192043 | 00:58 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Add jsonschema validation for creating secgroup https://review.openstack.org/189544 | 00:58 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Remove unused "id" and "rules" from secgroup body https://review.openstack.org/192029 | 00:58 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Fix a request body template for secgroup tests https://review.openstack.org/192031 | 00:58 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Fix test_create_security_group_with_no_name https://review.openstack.org/192030 | 00:59 |
| *** dsanders has quit IRC | 00:59 | |
| *** bradjones has joined #openstack-nova | 00:59 | |
| *** bradjones has quit IRC | 00:59 | |
| *** bradjones has joined #openstack-nova | 00:59 | |
| *** aloga has quit IRC | 01:00 | |
| *** dboik has joined #openstack-nova | 01:02 | |
| openstackgerrit | Zhenzan Zhou proposed openstack/nova: Bypass ironic server not available issue https://review.openstack.org/173681 | 01:04 |
| *** bradjones has quit IRC | 01:07 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/nova-specs: Ironic: Multiple compute host support https://review.openstack.org/194453 | 01:07 |
| *** dboik has quit IRC | 01:07 | |
| *** bradjones has joined #openstack-nova | 01:09 | |
| *** bradjones has quit IRC | 01:09 | |
| *** bradjones has joined #openstack-nova | 01:09 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Kicking the oslo.service tires https://review.openstack.org/192900 | 01:12 |
| *** Longgeek has joined #openstack-nova | 01:15 | |
| *** jasondotstar has joined #openstack-nova | 01:15 | |
| *** Kennan has quit IRC | 01:16 | |
| *** Kennan has joined #openstack-nova | 01:16 | |
| *** nic has quit IRC | 01:19 | |
| *** nic has joined #openstack-nova | 01:22 | |
| *** cing has joined #openstack-nova | 01:23 | |
| *** patrickeast has quit IRC | 01:23 | |
| *** wwriverrat has quit IRC | 01:24 | |
| *** sdake has joined #openstack-nova | 01:24 | |
| *** sdake_ has joined #openstack-nova | 01:27 | |
| *** aloga has joined #openstack-nova | 01:29 | |
| *** sdake has quit IRC | 01:31 | |
| openstackgerrit | Chung Chih, Hung proposed openstack/nova: Use monotonic.monotonic and stopwatches instead of time.time https://review.openstack.org/194073 | 01:31 |
| *** yamahata has quit IRC | 01:33 | |
| openstackgerrit | Chung Chih, Hung proposed openstack/nova: Use monotonic.monotonic and stopwatches instead of time.time https://review.openstack.org/194073 | 01:34 |
| *** alexpilotti has quit IRC | 01:40 | |
| *** zzzeek has quit IRC | 01:43 | |
| *** otter768 has joined #openstack-nova | 01:44 | |
| *** asettle has quit IRC | 01:44 | |
| *** dramakri has quit IRC | 01:44 | |
| openstackgerrit | Rui Chen proposed openstack/nova-specs: CreateVM supports subnet specified https://review.openstack.org/169638 | 01:45 |
| *** nic has quit IRC | 01:45 | |
| *** IanGovett has quit IRC | 01:48 | |
| *** otter768 has quit IRC | 01:49 | |
| *** LukeHinds has quit IRC | 01:52 | |
| *** park_heijlong has joined #openstack-nova | 01:55 | |
| *** sdake_ has quit IRC | 01:58 | |
| *** sdake has joined #openstack-nova | 01:58 | |
| *** unicell1 has quit IRC | 02:02 | |
| *** liusheng has joined #openstack-nova | 02:02 | |
| *** claudiub has quit IRC | 02:04 | |
| *** cfriesen_ has joined #openstack-nova | 02:06 | |
| *** sdake has quit IRC | 02:07 | |
| *** tjones1 has quit IRC | 02:10 | |
| *** Marga_ has joined #openstack-nova | 02:12 | |
| *** patrickeast has joined #openstack-nova | 02:14 | |
| *** hanzhang has joined #openstack-nova | 02:17 | |
| *** zzzeek has joined #openstack-nova | 02:17 | |
| *** sorrison_laptop has quit IRC | 02:20 | |
| *** sorrison_laptop has joined #openstack-nova | 02:21 | |
| *** sorrison_laptop has quit IRC | 02:22 | |
| *** yingjun has quit IRC | 02:25 | |
| *** ctrath has joined #openstack-nova | 02:26 | |
| *** adrian_otto has joined #openstack-nova | 02:28 | |
| *** adrian_otto1 has joined #openstack-nova | 02:29 | |
| *** bkopilov has quit IRC | 02:29 | |
| *** yingjun has joined #openstack-nova | 02:30 | |
| *** lyarwood has quit IRC | 02:32 | |
| *** lyarwood has joined #openstack-nova | 02:32 | |
| *** adrian_otto has quit IRC | 02:33 | |
| *** adrian_otto1 has quit IRC | 02:33 | |
| *** dims has quit IRC | 02:36 | |
| *** zhenguo has joined #openstack-nova | 02:40 | |
| *** zzzeek has quit IRC | 02:41 | |
| *** annashen has joined #openstack-nova | 02:42 | |
| *** ctrath has quit IRC | 02:42 | |
| *** pixelb has quit IRC | 02:42 | |
| *** pixelb has joined #openstack-nova | 02:42 | |
| *** linkedinyou has joined #openstack-nova | 02:46 | |
| *** dave-mccowan has joined #openstack-nova | 02:47 | |
| *** jasondotstar has quit IRC | 02:47 | |
| *** yingjun has quit IRC | 02:49 | |
| *** asettle has joined #openstack-nova | 02:51 | |
| *** ctrath has joined #openstack-nova | 02:52 | |
| *** rushiagr_away is now known as rushiagr | 02:53 | |
| *** Longgeek has quit IRC | 02:55 | |
| *** vilobhmm has quit IRC | 02:55 | |
| *** yingjun has joined #openstack-nova | 02:56 | |
| *** ctrath has quit IRC | 03:08 | |
| *** Nikolay_St has joined #openstack-nova | 03:09 | |
| *** ijw_ has quit IRC | 03:10 | |
| *** ijw_ has joined #openstack-nova | 03:11 | |
| *** dsanders has joined #openstack-nova | 03:12 | |
| *** Nikolay_St has quit IRC | 03:15 | |
| *** ijw_ has quit IRC | 03:16 | |
| *** electrocucaracha has joined #openstack-nova | 03:17 | |
| *** annashen has quit IRC | 03:19 | |
| openstackgerrit | shihanzhang proposed openstack/nova: Optimize validate_networks https://review.openstack.org/175742 | 03:19 |
| *** erlon has quit IRC | 03:20 | |
| *** jogo has joined #openstack-nova | 03:22 | |
| *** garyk has quit IRC | 03:22 | |
| *** wwriverrat has joined #openstack-nova | 03:22 | |
| *** Longgeek has joined #openstack-nova | 03:23 | |
| *** achanda has joined #openstack-nova | 03:25 | |
| *** sdake has joined #openstack-nova | 03:25 | |
| *** Marga_ has quit IRC | 03:26 | |
| *** tobe has joined #openstack-nova | 03:26 | |
| *** sdake_ has joined #openstack-nova | 03:27 | |
| *** electrocucaracha has quit IRC | 03:30 | |
| *** electrocucaracha has joined #openstack-nova | 03:30 | |
| *** sdake has quit IRC | 03:31 | |
| *** sdake__ has joined #openstack-nova | 03:32 | |
| *** rushiagr is now known as rushiagr_away | 03:32 | |
| *** dsanders has quit IRC | 03:32 | |
| *** dave-mccowan has quit IRC | 03:33 | |
| *** mgarza has joined #openstack-nova | 03:35 | |
| *** sdake_ has quit IRC | 03:35 | |
| openstackgerrit | Fei Long Wang proposed openstack/nova: Refactor image's code to support both versions https://review.openstack.org/144875 | 03:38 |
| *** achanda has quit IRC | 03:39 | |
| *** harlowja has quit IRC | 03:43 | |
| *** yingjun has quit IRC | 03:44 | |
| *** otter768 has joined #openstack-nova | 03:45 | |
| *** coolsvap|away is now known as coolsvap | 03:47 | |
| *** bkopilov has joined #openstack-nova | 03:47 | |
| *** yingjun has joined #openstack-nova | 03:48 | |
| *** avishay_ has joined #openstack-nova | 03:49 | |
| *** rajesht has joined #openstack-nova | 03:50 | |
| *** otter768 has quit IRC | 03:50 | |
| *** garyk has joined #openstack-nova | 03:51 | |
| *** ajayaa has joined #openstack-nova | 03:54 | |
| *** vilobhmm has joined #openstack-nova | 03:56 | |
| *** yingjun has quit IRC | 03:56 | |
| *** tojuvone has joined #openstack-nova | 03:57 | |
| *** mgarza has quit IRC | 03:58 | |
| *** iamjarvo has joined #openstack-nova | 03:59 | |
| *** brad[] has joined #openstack-nova | 03:59 | |
| *** dramakri has joined #openstack-nova | 03:59 | |
| *** dramakri has left #openstack-nova | 04:00 | |
| *** Longgeek has quit IRC | 04:00 | |
| *** Longgeek has joined #openstack-nova | 04:00 | |
| *** kiran-r has joined #openstack-nova | 04:01 | |
| *** pixelb has quit IRC | 04:06 | |
| *** dsanders has joined #openstack-nova | 04:09 | |
| *** unicell has joined #openstack-nova | 04:09 | |
| *** rushiagr_away is now known as rushiagr | 04:11 | |
| *** Marga_ has joined #openstack-nova | 04:14 | |
| *** rushiagr is now known as rushiagr_away | 04:14 | |
| *** mtanino has quit IRC | 04:15 | |
| *** rushiagr_away is now known as rushiagr | 04:15 | |
| *** jogo has quit IRC | 04:16 | |
| *** mestery has joined #openstack-nova | 04:19 | |
| *** kaisers has quit IRC | 04:19 | |
| *** jogo has joined #openstack-nova | 04:20 | |
| *** kaisers has joined #openstack-nova | 04:22 | |
| *** sdake__ has quit IRC | 04:24 | |
| *** ildikov has quit IRC | 04:24 | |
| *** iamjarvo has quit IRC | 04:25 | |
| *** dsanders has quit IRC | 04:27 | |
| openstackgerrit | Zhang Ni proposed openstack/nova-specs: Improve nova attach volume api to support disk-bus https://review.openstack.org/185973 | 04:28 |
| *** Longgeek has quit IRC | 04:29 | |
| *** Marga_ has quit IRC | 04:31 | |
| *** cfriesen_ has quit IRC | 04:31 | |
| openstackgerrit | Zhang Ni proposed openstack/nova-specs: Add volume type to create server API https://review.openstack.org/187812 | 04:32 |
| *** Longgeek has joined #openstack-nova | 04:32 | |
| *** electrocucaracha has quit IRC | 04:33 | |
| *** brad[] has quit IRC | 04:35 | |
| *** Marga_ has joined #openstack-nova | 04:36 | |
| *** lyarwood has quit IRC | 04:37 | |
| *** lyarwood has joined #openstack-nova | 04:38 | |
| *** Marga_ has quit IRC | 04:38 | |
| *** mgarza_ has joined #openstack-nova | 04:38 | |
| *** Marga_ has joined #openstack-nova | 04:38 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: convert driver to use nova.objects.ImageMeta https://review.openstack.org/189085 | 04:39 |
| *** husanu has quit IRC | 04:39 | |
| *** Nikolay_St has joined #openstack-nova | 04:40 | |
| openstackgerrit | Zhang Ni proposed openstack/nova-specs: Improve nova attach volume api to support disk-bus https://review.openstack.org/185973 | 04:41 |
| *** mgarza__ has joined #openstack-nova | 04:41 | |
| *** husanu has joined #openstack-nova | 04:41 | |
| *** deepthi has joined #openstack-nova | 04:42 | |
| *** mgarza_ has quit IRC | 04:43 | |
| *** arnaud___ has joined #openstack-nova | 04:43 | |
| *** linkedinyou has quit IRC | 04:49 | |
| *** mestery has quit IRC | 04:49 | |
| *** husanu has quit IRC | 04:50 | |
| *** takedakn has joined #openstack-nova | 04:51 | |
| *** husanu has joined #openstack-nova | 04:52 | |
| *** brad[] has joined #openstack-nova | 04:54 | |
| *** _gryf|afk is now known as _gryf | 04:58 | |
| *** dsanders has joined #openstack-nova | 04:59 | |
| *** asettle has quit IRC | 05:00 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: add support for cores per socket https://review.openstack.org/187942 | 05:01 |
| *** husanu has quit IRC | 05:01 | |
| *** takedakn has quit IRC | 05:01 | |
| *** unicell1 has joined #openstack-nova | 05:02 | |
| *** husanu has joined #openstack-nova | 05:03 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: create common object for limits, reservations and shares https://review.openstack.org/179059 | 05:03 |
| *** nkrinner has joined #openstack-nova | 05:04 | |
| *** unicell has quit IRC | 05:04 | |
| *** krtaylor has quit IRC | 05:05 | |
| *** armax has joined #openstack-nova | 05:08 | |
| *** tobe has quit IRC | 05:11 | |
| *** boris-42 has quit IRC | 05:12 | |
| *** Nikolay_St has quit IRC | 05:13 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: Resource limits for memory https://review.openstack.org/179060 | 05:14 |
| *** jmccrory has quit IRC | 05:15 | |
| *** jmccrory has joined #openstack-nova | 05:16 | |
| *** yingjun has joined #openstack-nova | 05:16 | |
| *** tobe has joined #openstack-nova | 05:17 | |
| *** Marga_ has quit IRC | 05:17 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: store extra_specs object https://review.openstack.org/179061 | 05:18 |
| openstackgerrit | garyk proposed openstack/nova: VMware: add resource limits for disk https://review.openstack.org/179062 | 05:18 |
| *** isd has joined #openstack-nova | 05:18 | |
| *** tobe has quit IRC | 05:19 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: add in vnic resource limitations https://review.openstack.org/180202 | 05:23 |
| *** patrickeast has quit IRC | 05:23 | |
| *** patrickeast has joined #openstack-nova | 05:24 | |
| *** krtaylor has joined #openstack-nova | 05:25 | |
| *** irenab has joined #openstack-nova | 05:25 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: support limits via image meta data https://review.openstack.org/180942 | 05:26 |
| *** lyarwood has quit IRC | 05:31 | |
| *** lyarwood has joined #openstack-nova | 05:31 | |
| openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Support --all-tenants in floatingips-list https://review.openstack.org/171389 | 05:35 |
| *** fawadkhaliq has joined #openstack-nova | 05:38 | |
| *** ildikov has joined #openstack-nova | 05:41 | |
| *** arnaud___ has quit IRC | 05:42 | |
| *** arnaud___ has joined #openstack-nova | 05:44 | |
| *** otter768 has joined #openstack-nova | 05:46 | |
| *** sudipto has joined #openstack-nova | 05:48 | |
| *** arnaud___ has quit IRC | 05:48 | |
| *** otter768 has quit IRC | 05:51 | |
| *** salv-orlando has joined #openstack-nova | 05:54 | |
| *** patrickeast has quit IRC | 05:56 | |
| *** husanu has quit IRC | 05:56 | |
| *** salv-orlando has quit IRC | 05:56 | |
| *** doron_afk has joined #openstack-nova | 05:57 | |
| *** avishay_ is now known as avishay | 05:57 | |
| *** vilobhmm has quit IRC | 05:58 | |
| *** salv-orlando has joined #openstack-nova | 05:58 | |
| *** husanu has joined #openstack-nova | 05:58 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Use host AZ instead of instance AZ in scheduler request_spec https://review.openstack.org/174269 | 06:01 |
| *** annashen has joined #openstack-nova | 06:01 | |
| *** radek__ has joined #openstack-nova | 06:02 | |
| *** exploreshaifali has joined #openstack-nova | 06:02 | |
| *** yingjun has quit IRC | 06:04 | |
| *** radek__ has quit IRC | 06:06 | |
| *** radek__ has joined #openstack-nova | 06:07 | |
| *** shihanzhang has joined #openstack-nova | 06:11 | |
| *** Kennan has quit IRC | 06:11 | |
| *** Kennan2 has joined #openstack-nova | 06:11 | |
| *** oro has joined #openstack-nova | 06:13 | |
| *** hanzhang has quit IRC | 06:14 | |
| *** isd has quit IRC | 06:16 | |
| *** mgarza__ has quit IRC | 06:16 | |
| *** browne has quit IRC | 06:17 | |
| *** lpetrut has joined #openstack-nova | 06:18 | |
| *** yingjun has joined #openstack-nova | 06:19 | |
| *** nihilifer has joined #openstack-nova | 06:21 | |
| *** tobe has joined #openstack-nova | 06:22 | |
| *** ssurana has joined #openstack-nova | 06:23 | |
| *** ssurana has quit IRC | 06:25 | |
| *** belmoreira has joined #openstack-nova | 06:26 | |
| *** ssurana has joined #openstack-nova | 06:27 | |
| *** dsanders has quit IRC | 06:29 | |
| *** wanghua has joined #openstack-nova | 06:32 | |
| openstackgerrit | Hiroyuki Eguchi proposed openstack/nova: qemu-img convert should be skipped when migrating https://review.openstack.org/194516 | 06:34 |
| *** ianbrown has quit IRC | 06:38 | |
| *** ssurana has quit IRC | 06:38 | |
| *** shihanzhang has quit IRC | 06:39 | |
| *** lyarwood has quit IRC | 06:39 | |
| *** ndipanov has joined #openstack-nova | 06:40 | |
| *** markvoelker has quit IRC | 06:40 | |
| *** lyarwood has joined #openstack-nova | 06:41 | |
| *** rushiagr is now known as rushiagr_away | 06:42 | |
| *** exploreshaifali has quit IRC | 06:43 | |
| openstackgerrit | Hans Lindgren proposed openstack/nova: Remove/deprecate unused parts of the compute node object https://review.openstack.org/186757 | 06:44 |
| openstackgerrit | Hans Lindgren proposed openstack/nova: Make ComputeNode.service_id nullable to match db schema https://review.openstack.org/169284 | 06:44 |
| *** deepthi has quit IRC | 06:44 | |
| openstackgerrit | Abhishek Kekane proposed openstack/nova: libvirt: Kill rsync/scp processes before deleting instance https://review.openstack.org/192986 | 06:45 |
| openstackgerrit | Hans Lindgren proposed openstack/nova: Avoid querying for Service in resource tracker https://review.openstack.org/169288 | 06:46 |
| *** ndipanov has quit IRC | 06:47 | |
| *** hanzhang has joined #openstack-nova | 06:49 | |
| *** armax has quit IRC | 06:49 | |
| *** sahid has joined #openstack-nova | 06:51 | |
| *** mmedvede has quit IRC | 06:53 | |
| *** Kennan2 is now known as Kennan | 06:53 | |
| *** dsanders has joined #openstack-nova | 06:53 | |
| *** fawadkhaliq has quit IRC | 06:55 | |
| *** mpaolino has joined #openstack-nova | 06:55 | |
| *** fawadkhaliq has joined #openstack-nova | 06:55 | |
| *** Longgeek_ has joined #openstack-nova | 06:56 | |
| *** Nikolay_St has joined #openstack-nova | 06:56 | |
| *** abhishekk has joined #openstack-nova | 06:57 | |
| *** sahid has quit IRC | 06:59 | |
| *** sahid has joined #openstack-nova | 06:59 | |
| *** Longgeek has quit IRC | 06:59 | |
| openstackgerrit | Rajesh Tailor proposed openstack/nova: Remove unnecessary ValueError exception https://review.openstack.org/190525 | 07:01 |
| *** lyarwood has quit IRC | 07:03 | |
| *** doron_afk is now known as doron_ | 07:03 | |
| *** lyarwood has joined #openstack-nova | 07:03 | |
| *** avishay_ has joined #openstack-nova | 07:04 | |
| *** avishay has quit IRC | 07:05 | |
| *** bkopilov has quit IRC | 07:05 | |
| *** dsanders has quit IRC | 07:05 | |
| *** RuiChen has joined #openstack-nova | 07:06 | |
| *** Jabadia has joined #openstack-nova | 07:07 | |
| *** park_heijlong has quit IRC | 07:08 | |
| *** lyarwood has quit IRC | 07:09 | |
| *** lyarwood has joined #openstack-nova | 07:09 | |
| *** jlanoux has joined #openstack-nova | 07:10 | |
| *** matrohon has joined #openstack-nova | 07:11 | |
| *** vilobhmm has joined #openstack-nova | 07:13 | |
| *** dsanders has joined #openstack-nova | 07:16 | |
| *** annashen has quit IRC | 07:16 | |
| *** avishay_ has quit IRC | 07:18 | |
| *** park_heijlong has joined #openstack-nova | 07:19 | |
| *** moshele has joined #openstack-nova | 07:20 | |
| *** doron_ is now known as doron_afk | 07:20 | |
| *** haypo has joined #openstack-nova | 07:20 | |
| *** haypo has quit IRC | 07:20 | |
| *** haypo has joined #openstack-nova | 07:20 | |
| *** matrohon has quit IRC | 07:20 | |
| openstackgerrit | PrashantUpadhyay proposed openstack/nova: check for volume size https://review.openstack.org/194523 | 07:20 |
| bauzas | tonyb: around ? | 07:21 |
| *** prashup has joined #openstack-nova | 07:21 | |
| *** bkopilov has joined #openstack-nova | 07:21 | |
| tonyb | bauzas: Yup | 07:21 |
| tonyb | bauzas: but not for too long :( | 07:21 |
| tonyb | bauzas: I assume this is related to my email ;P | 07:21 |
| *** oro has quit IRC | 07:22 | |
| bauzas | tonyb: coolio, Australians and Frenchs can be discussing not only for the Rugby World Cup then | 07:22 |
| tonyb | bauzas: :) | 07:22 |
| openstackgerrit | wangxiyuan proposed openstack/nova: Remove an useless variable and fix a typo in api https://review.openstack.org/187118 | 07:22 |
| bauzas | tonyb: so, pinging you because it's public, right? | 07:22 |
| bauzas | tonyb: I mean, okay for discussing here? | 07:22 |
| tonyb | bauzas: Yup all public, but we'd still like some attention / validation | 07:23 |
| bauzas | tonyb: okay | 07:23 |
| sudipto | johnthetubaguy, Hi, https://review.openstack.org/#/c/192760/ in this review, dansmith mentions that we should not actually drop the table column from the table and wanted me to create a whitelist of attributes that need cleanup in the test_migrations.py - to have a list of attributes that can be cleaned up from the DB in the 'M' release. Your thoughts? | 07:23 |
| bauzas | tonyb: so https://bugs.launchpad.net/nova/+bug/1456228 | 07:23 |
| openstack | Launchpad bug 1456228 in OpenStack Security Advisory "Trusted vm can be powered on untrusted host" [Undecided,Incomplete] | 07:23 |
| bauzas | tonyb: I saw that bug a couple of times before | 07:24 |
| tonyb | bauzas: okay. good to know | 07:24 |
| *** romainh has joined #openstack-nova | 07:24 | |
| tonyb | bauzas: So the basic outline of the bug sounds feasible but I don't have the hardware nor the expertise to triage it | 07:25 |
| *** wwriverrat has quit IRC | 07:25 | |
| tonyb | bauzas: I suspect it's not a nova bug. but IF the modified server does update the Attestation service then it would be a nova issue | 07:26 |
| tonyb | bauzas: but really the nova code is just an interface to that service (IIUC) | 07:26 |
| openstackgerrit | heijlong proposed openstack/nova: Merge used-limits functional tests of v2 and v2.1 https://review.openstack.org/174215 | 07:28 |
| *** lpetrut has quit IRC | 07:29 | |
| bauzas | tonyb: sorry I'm otp | 07:31 |
| bauzas | tonyb: still there in 10 mins ? | 07:31 |
| tonyb | bauzas: 5? | 07:31 |
| bauzas | tonyb: trying | 07:32 |
| *** ijuwang_away has quit IRC | 07:32 | |
| tonyb | bauzas: Sorry to be a pain but my wife's holding dinner for me. | 07:32 |
| tonyb | bauzas: I can get back online in say 90mins and we can chat longer if that's a better option? | 07:32 |
| *** boris-42 has joined #openstack-nova | 07:32 | |
| *** lucasagomes has joined #openstack-nova | 07:32 | |
| *** ijuwang_away has joined #openstack-nova | 07:32 | |
| *** avishay_ has joined #openstack-nova | 07:32 | |
| tonyb | bauzas: Sorry I'm going to have to bail. I'll be back when the kids are in bed about 90mins from now | 07:36 |
| tonyb | bauzas: sorry :( | 07:36 |
| bauzas | tonyb: okay, sorry about that | 07:37 |
| *** salv-orl_ has joined #openstack-nova | 07:37 | |
| bauzas | tonyb: so, ping me if you have time later on, we can discuss on the TrustedFilter | 07:37 |
| bauzas | tonyb: tbc, my opinion is that the TrustedFilter should be landed in hell (or out-of-tree) | 07:38 |
| bauzas | tonyb: because that's the only in-tree filter which calls an API for saying OK or not | 07:38 |
| *** fawadkhaliq has quit IRC | 07:39 | |
| *** salv-orlando has quit IRC | 07:40 | |
| bauzas | tonyb: in the meantime, I'll provide some insights on the bug | 07:40 |
| *** markvoelker has joined #openstack-nova | 07:41 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Use host AZ instead of instance AZ in availability zone filter https://review.openstack.org/174269 | 07:42 |
| openstackgerrit | Qin Zhao proposed openstack/nova: Consolidate code to get the correct availability zone of an instance https://review.openstack.org/149562 | 07:42 |
| *** mpavone has joined #openstack-nova | 07:43 | |
| *** fawadkhaliq has joined #openstack-nova | 07:44 | |
| *** markvoelker has quit IRC | 07:46 | |
| *** linkedinyou has joined #openstack-nova | 07:46 | |
| *** otter768 has joined #openstack-nova | 07:47 | |
| *** markus_z has joined #openstack-nova | 07:47 | |
| *** dsanders has quit IRC | 07:47 | |
| *** sahid has quit IRC | 07:49 | |
| *** sahid_ has joined #openstack-nova | 07:49 | |
| *** hanzhang has quit IRC | 07:49 | |
| *** hanzhang has joined #openstack-nova | 07:50 | |
| *** claudiub has joined #openstack-nova | 07:51 | |
| *** otter768 has quit IRC | 07:51 | |
| *** masayukig has quit IRC | 07:52 | |
| *** linkedinyou has quit IRC | 07:53 | |
| *** masayukig has joined #openstack-nova | 07:54 | |
| *** deepthi has joined #openstack-nova | 07:55 | |
| *** cjwei has quit IRC | 07:55 | |
| *** avishay_ has quit IRC | 07:56 | |
| *** avishay__ has joined #openstack-nova | 07:56 | |
| openstackgerrit | Andreas Scheuring proposed openstack/nova-specs: Spec to Add 'macvtap' as vif type to novas libvirt driver. https://review.openstack.org/182280 | 07:57 |
| *** kiran-r has quit IRC | 07:57 | |
| *** salv-orl_ has quit IRC | 07:58 | |
| *** LukeHinds has joined #openstack-nova | 07:58 | |
| *** sahid_ has quit IRC | 07:59 | |
| *** sahid has joined #openstack-nova | 07:59 | |
| *** salv-orlando has joined #openstack-nova | 08:00 | |
| *** claudiub has quit IRC | 08:00 | |
| *** mpaolino has quit IRC | 08:01 | |
| *** mpaolino has joined #openstack-nova | 08:03 | |
| *** chen12 has joined #openstack-nova | 08:04 | |
| *** karimb has joined #openstack-nova | 08:04 | |
| chen12 | hello.I get an error for nova when install devstack,anyone know how to fix this : http://paste.openstack.org/show/315959/ | 08:04 |
| *** claudiub has joined #openstack-nova | 08:05 | |
| *** claudiub|2 has joined #openstack-nova | 08:07 | |
| lifeless | chen12: you'remissing some tags from the openstack repo | 08:08 |
| chen12 | lifeless, how to add these tags ?? I already run "git pull origin master" | 08:09 |
| lifeless | lets check | 08:09 |
| lifeless | git tag | 08:09 |
| chen12 | lifeless, and I have run clean.sh & stack.sh in devstack.... | 08:10 |
| chen12 | lifeless, http://paste.openstack.org/show/315960/ | 08:10 |
| *** vilobhmm has quit IRC | 08:10 | |
| *** claudiub has quit IRC | 08:10 | |
| openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova: scheduler: weight nodes based on their cached images https://review.openstack.org/173304 | 08:11 |
| openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova: Add new image cache monitor in compute nodes https://review.openstack.org/173303 | 08:11 |
| openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova: vmware: add image_cache_manager property https://review.openstack.org/187602 | 08:11 |
| openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova: compute monitors: set update delta from base class https://review.openstack.org/173987 | 08:11 |
| lifeless | yeah you're missing 12.0.0a0 | 08:11 |
| lifeless | what's your git remote for origin ? | 08:12 |
| johnthetubaguy | sudipto: yeah, I think he has a good point, although we totally need to get those rules written up. I think he added some new rules recently | 08:12 |
| markus_z | chen12: You could try "RECLONE=yes" in the "local.conf" file. Or in the existing nova repository "git fetch --tags". | 08:13 |
| *** claudiub has joined #openstack-nova | 08:14 | |
| johnthetubaguy | bauzas: ah, thats the last filter that does nasty things, interesting, do we give them a good way to do it with meters, or something, instead? | 08:14 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Added server tags support in nova-api https://review.openstack.org/128940 | 08:14 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Added method exists to the Tag object https://review.openstack.org/178103 | 08:14 |
| bauzas | johnthetubaguy: the problem with that filter is that it's checking something out of Nova | 08:14 |
| bauzas | johnthetubaguy: the other filters are checking the HostState only | 08:15 |
| johnthetubaguy | bauzas: agreed | 08:15 |
| johnthetubaguy | bauzas: I am saying, did we give them a good way to get that data into the HostState | 08:15 |
| bauzas | johnthetubaguy: well, I'm usually -1 for some metrics not provided by the virt driver | 08:15 |
| johnthetubaguy | bauzas: sounds a bit like the metric/meetering stuff might do that | 08:15 |
| *** yingjun has quit IRC | 08:15 | |
| johnthetubaguy | bauzas: well the virt driver can call out to the service and report those stats, although they may dislike that approach | 08:16 |
| bauzas | johnthetubaguy: the rule is that we're providing the metrics returned by the virt driver | 08:16 |
| *** claudiub|2 has quit IRC | 08:16 | |
| johnthetubaguy | bauzas: yes, I am saying make the virt driver return that info | 08:16 |
| bauzas | johnthetubaguy: so, given that, perhaps something running on the node which would be called by the virt driver would be OK | 08:16 |
| bauzas | johnthetubaguy: lemme think about that | 08:16 |
| bauzas | johnthetubaguy: the main problem I have is that I dunno how the Attestation API is returning True or False | 08:17 |
| johnthetubaguy | bauzas: its a horrible thing to do, but I love the use case | 08:17 |
| chen12 | lifeless, origin https://git.openstack.org/openstack/nova.git (fetch) & origin https://git.openstack.org/openstack/nova.git (push) | 08:17 |
| chen12 | markus_z, lifeless "git fetch --tags" works | 08:18 |
| lifeless | cool | 08:18 |
| lifeless | wonder why it was missing in the first place though | 08:18 |
| lifeless | this isn't the first report | 08:18 |
| lifeless | I think I'll ask infra to check the mirrors | 08:18 |
| bauzas | johnthetubaguy: what I need to understand is why we need an Attestation API, if that Attestation API is opensource or not | 08:18 |
| *** avishay has joined #openstack-nova | 08:18 | |
| bauzas | johnthetubaguy: and see how we can have the drivers using it or not | 08:18 |
| markus_z | chen12: lifeless: It could be a new execution of "stack.sh" on an existing devstack installation. | 08:18 |
| markus_z | The tag was introduced 5 days ago. | 08:19 |
| markus_z | https://git.openstack.org/cgit/openstack/nova/refs/ | 08:19 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Removed extra '-' from rest_api_version_history.rst https://review.openstack.org/194536 | 08:19 |
| bauzas | johnthetubaguy: but you know, it would be like the same problem as raised by mikal here http://lists.openstack.org/pipermail/openstack-dev/2015-June/067630.html | 08:19 |
| bauzas | johnthetubaguy: which means that we would have an in-tree dependency on an external tool | 08:19 |
| johnthetubaguy | bauzas: so I thought it was a rest API in this case, I guess its not | 08:20 |
| lifeless | markus_z: no, it can't be that | 08:20 |
| johnthetubaguy | bauzas: but yes, we have that problem all over the place, sadly :( | 08:20 |
| *** avishay__ has quit IRC | 08:20 | |
| lifeless | markus_z: because the error indicates that the new master branch is present | 08:20 |
| lifeless | markus_z: which means a git pull was done | 08:20 |
| bauzas | johnthetubaguy: I think the idea is, we use the nova-scheduler as something we know uncompromised | 08:20 |
| johnthetubaguy | bauzas: I agree | 08:21 |
| bauzas | johnthetubaguy: because using nova-compute for checking its own state if it's compromised is kinda weird | 08:21 |
| johnthetubaguy | bauzas: I am saying the virt driver calls the API and reports the stat to the scheduler | 08:21 |
| garyk | johnthetubaguy: can you please assign the bp https://blueprints.launchpad.net/openstack/?searchtext=cells-instance-migration to me. i have spoke with alaski about this and have posted an initial implementation | 08:21 |
| johnthetubaguy | bauzas: true, although its more about the hypervisor than the nova-compute VM, as I understood it | 08:21 |
| lifeless | markus_z: unless devstack does omething really weird? | 08:21 |
| bauzas | johnthetubaguy: but since the virt thing is running on something potentially compromised, how can we be sure that what's reported by that virt driver is safe ? | 08:21 |
| lifeless | markus_z: 20:21 < clarkb> anyone can check, http://git0X.openstack.org:8080/cgit/openstack/nova looks like 01 is missing it | 08:22 |
| lifeless | markus_z: from -infra. | 08:22 |
| bauzas | johnthetubaguy: the Attestation API is not checking the VM, rather the host | 08:22 |
| johnthetubaguy | garyk: did you add the blueprint tag to your patches? I don't see them on the whiteboard of the blueprint | 08:22 |
| bauzas | johnthetubaguy: IIUC again, since I have zero understanding on which software is giving us that | 08:22 |
| garyk | johnthetubaguy: yes | 08:22 |
| johnthetubaguy | bauzas: yes, its checking the host, understood | 08:22 |
| *** yassine_ has joined #openstack-nova | 08:22 | |
| garyk | johnthetubaguy: at the moment it is one patch. not sure i understand | 08:22 |
| johnthetubaguy | garyk: my bad, I thought that was the spec for some reason | 08:23 |
| garyk | johnthetubaguy: you gotta get out the spec weeds and back into the trenches :) | 08:23 |
| johnthetubaguy | garyk: cool, that should be assigned to you know, I think | 08:23 |
| markus_z | lifeless: thanks for the info, this will safe me some time | 08:23 |
| chen12 | markus_z, lifeless it not just happened for nova, my horizon has the same issue too. | 08:24 |
| johnthetubaguy | garyk: code, what is that? | 08:24 |
| lifeless | chen12: ok, so do a fetch --tags, then use te git0X urls to figure out which ones are msising it and report in -infra | 08:24 |
| garyk | my thinking is that it is something that is interpretted - in the past compiled | 08:24 |
| *** alex_klimov has joined #openstack-nova | 08:25 | |
| *** danpb has joined #openstack-nova | 08:25 | |
| *** yingjun has joined #openstack-nova | 08:25 | |
| chen12 | lifeless, what is git0X ? =>> Your requested host "git0x.openstack.org" could not be resolved by DNS. | 08:26 |
| lifeless | chen12: X should be a number :) | 08:26 |
| bauzas | johnthetubaguy: okay, lemme open a thread for that, asking for more details | 08:26 |
| johnthetubaguy | bauzas: you are right though, the host claim-ing its not comprimised is odd | 08:27 |
| openstackgerrit | Deepak C Shetty (deepakcs) proposed openstack/nova: Nova part of volume snapshot improvements https://review.openstack.org/172813 | 08:27 |
| openstackgerrit | Deepak C Shetty (deepakcs) proposed openstack/nova: objects: Add support for OnlineSnapshotStatus object https://review.openstack.org/194538 | 08:27 |
| johnthetubaguy | bauzas: I think this go back to the idea of adding an API to add to the host stats | 08:27 |
| *** deepthi has quit IRC | 08:27 | |
| bauzas | johnthetubaguy: what I would appreciate is a logic invert where https://github.com/OpenAttestation/OpenAttestation would populate its views to the Scheduler metrics | 08:27 |
| johnthetubaguy | so some intermediate service periodically updates the host stats via the scheduler REST API (yes, the thing that doesn't exist yet) | 08:27 |
| *** Marga_ has joined #openstack-nova | 08:28 | |
| openstackgerrit | Tang Chen proposed openstack/nova: Add realize_host_maintenance() API to conductor. https://review.openstack.org/194539 | 08:28 |
| openstackgerrit | Tang Chen proposed openstack/nova: Add realize_host_maintenance() API to compute. https://review.openstack.org/194540 | 08:28 |
| openstackgerrit | Tang Chen proposed openstack/nova: Enable synchronous live migration in libvirt driver. https://review.openstack.org/194541 | 08:28 |
| openstackgerrit | Tang Chen proposed openstack/nova: Enable host maintenance mode in libvirt driver. https://review.openstack.org/194542 | 08:28 |
| openstackgerrit | Tang Chen proposed openstack/nova: Add unit tests for libvirt host maintenance mode. https://review.openstack.org/194543 | 08:28 |
| bauzas | johnthetubaguy: yeah that, like we did for aggregates and instances | 08:28 |
| *** Marga_ has quit IRC | 08:28 | |
| johnthetubaguy | bauzas: yeah, I think we are thinking along the same lines here | 08:28 |
| bauzas | johnthetubaguy: okay, lemme draft thru some bits | 08:28 |
| sudipto | johnthetubaguy, ok - so shall i just go ahead with the whitelist idea for now? | 08:28 |
| bauzas | johnthetubaguy: in the ML | 08:28 |
| *** Marga_ has joined #openstack-nova | 08:29 | |
| garyk | lifeless: you familiar with the script for checking the validity of requirements? | 08:30 |
| lifeless | garyk: intimiately | 08:31 |
| garyk | sweet. | 08:31 |
| lifeless | garyk: more salty, but sure. | 08:31 |
| garyk | quick question - in neutron there was the split out of drivers | 08:31 |
| lifeless | yes, I believe so. | 08:32 |
| garyk | in the test requirements we have: https://github.com/openstack/vmware-nsx/blob/master/test-requirements.txt#L5 | 08:32 |
| openstackgerrit | Ankit Agrawal proposed openstack/nova: Test patch libvirt race condition (do not merge) https://review.openstack.org/191732 | 08:32 |
| garyk | the script validation no barfs on that - http://logs.openstack.org/60/194360/1/check/gate-vmware-nsx-requirements/b294b53/console.html | 08:32 |
| lifeless | (sorry, bad humor :)) | 08:32 |
| garyk | is there anyway of having it skip that line? | 08:32 |
| garyk | similar to # noqa for pep8 | 08:33 |
| lifeless | well | 08:33 |
| lifeless | perhaps | 08:34 |
| lifeless | letme look | 08:34 |
| garyk | gracias | 08:34 |
| chen12 | lifeless, o.... Thanks! | 08:34 |
| lifeless | oh | 08:34 |
| lifeless | thats the one I haven't poked around inside much yet | 08:35 |
| lifeless | garyk: so, those lines aren't valid | 08:35 |
| lifeless | garyk: update-requirements has special logic to preserve them | 08:35 |
| park__heijlong | johnthetubaguy: could you help to continue to take a review about https://review.openstack.org/#/c/184201/, thank you in advance. | 08:35 |
| lifeless | garyk: but project-requirements-change.py which is newer as I understand it is more rigorous | 08:35 |
| *** linkedinyou has joined #openstack-nova | 08:35 | |
| johnthetubaguy | park__heijlong: its on my list | 08:36 |
| garyk | ok, i'll take a look… oof | 08:36 |
| *** linkedinyou has quit IRC | 08:36 | |
| park__heijlong | johnthetubaguy: sure, sorry for interrupt, thank you~ | 08:37 |
| lifeless | garyk: so, I've argued before that the neutron split-out here should just say 'neutron' in requirements.txt | 08:37 |
| lifeless | garyk: possibly version locked | 08:37 |
| garyk | i hear you. in this case we want to take the current master branch | 08:38 |
| lifeless | sure | 08:38 |
| lifeless | but those files end up in packaging metadata when you do sdist | 08:38 |
| lifeless | what you do for testing can and should be decoupled | 08:38 |
| garyk | that is, the branch lives in parallel to the master branch and when we cut a stable we freeze that one | 08:38 |
| lifeless | e.g. as a constraints file | 08:38 |
| *** moshele has quit IRC | 08:38 | |
| *** lpetrut has joined #openstack-nova | 08:39 | |
| lifeless | garyk: sure, thats orthogonal though | 08:40 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Removed extra '-' from rest_api_version_history.rst https://review.openstack.org/194536 | 08:42 |
| lifeless | garyk: that said, I know some of the bits are still very rough - e.g. pip install -c is only in git at the moment, not a release | 08:42 |
| garyk | ok, thanks. i am just not sure what we should do. we have the global bot posting the requiremenst but the check fails. what do you suggest? | 08:45 |
| *** prashup has quit IRC | 08:46 | |
| lifeless | well, it doesn't make sense that the two sides be inconsistent | 08:47 |
| lifeless | so you could put a patch together to the infra script to make it handle utls | 08:47 |
| lifeless | urls | 08:47 |
| garyk | lifeless: ok, will try that | 08:47 |
| lifeless | but - fair warning | 08:47 |
| lifeless | you may get pushback | 08:47 |
| lifeless | separately, you could examine why you need the url | 08:48 |
| *** lpetrut1 has joined #openstack-nova | 08:48 | |
| lifeless | like, what in your test setup is consuming it | 08:48 |
| *** lpetrut has quit IRC | 08:48 | |
| lifeless | and just move it somewhere else. e.g. tox.ini | 08:48 |
| *** rgerganov__ is now known as rgerganov | 08:48 | |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: devref: virtual machine states and transitions https://review.openstack.org/191021 | 08:48 |
| lifeless | for instance, removing it and putting it in requirements.txt (where it should be) as 'neutron', with an install_command of 'pip install -e https://....; pip install -r {deps} | 08:49 |
| lifeless | might work | 08:49 |
| *** arnaud____ has joined #openstack-nova | 08:49 | |
| *** prashup has joined #openstack-nova | 08:49 | |
| rgerganov | alex_xu_: gilliard: hi guys, any chance you can take a look at https://review.openstack.org/#/c/148509/ again? | 08:49 |
| *** moshele has joined #openstack-nova | 08:50 | |
| *** pixelb has joined #openstack-nova | 08:51 | |
| garyk | lifeless: do you by any chance have an example of that? | 08:52 |
| lifeless | garyk: nope | 08:52 |
| lifeless | garyk: sorry; only the discussions I've had with neutron folk about this :) | 08:52 |
| garyk | ok, thanks | 08:52 |
| *** aix has quit IRC | 08:53 | |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: devref: virtual machine states and transitions https://review.openstack.org/191021 | 08:54 |
| *** arnaud____ has quit IRC | 08:54 | |
| lifeless | garyk: but install_command is a tox thing | 08:54 |
| *** humble_ has joined #openstack-nova | 08:54 | |
| lifeless | garyk: so the idea is you'd install neutron and then the dep in requirements becomes a no-op in a testing context | 08:54 |
| lifeless | garyk: which is all you want anyway, right? | 08:54 |
| *** lyarwood has quit IRC | 08:55 | |
| lifeless | garyk: or you can use zuul-cloner | 08:55 |
| *** lyarwood has joined #openstack-nova | 08:55 | |
| lifeless | garyk: which AIUI you're really meant to since you want the ZUUL_REF | 08:55 |
| lifeless | garyk: if you don't have the ZUUL_REF being tested you can let bad things thread the needle | 08:55 |
| kashyap | markus_z: Hi, for the patch you just pushed, you might also want to add the rendered version (from staging, once available?) as a comment to see the output result. | 08:56 |
| lifeless | garyk: we should perhaps -> #openstack-neutron :) | 08:56 |
| markus_z | kashyap: sure, I'll do that | 08:56 |
| kashyap | markus_z: Thanks. | 08:57 |
| abhishekk | johnthetubaguy: hi, could you please review nova-specs https://review.openstack.org/135387 when you get time, thank you in advance. | 08:57 |
| *** wanghua has quit IRC | 08:58 | |
| chen12 | lifeless, should this ok ? https://storyboard.openstack.org/#!/story/2000300 | 08:58 |
| *** humble__ has joined #openstack-nova | 08:59 | |
| lifeless | chen12: seems reasonable; I have no idea if thats the rgiht place to file | 09:00 |
| lifeless | puppet-cgit is the logic for managing the cgit stuff, not operational maintenance of the mirrors | 09:01 |
| lifeless | chen12: I'd mention that in -infra and see if you get any bites | 09:01 |
| *** pkonisze has joined #openstack-nova | 09:01 | |
| *** pkonisze has quit IRC | 09:01 | |
| *** pkonisze has joined #openstack-nova | 09:01 | |
| lifeless | chen12: ah, you're not in #openstack-infra | 09:02 |
| *** avishay has quit IRC | 09:02 | |
| *** avishay_ has joined #openstack-nova | 09:02 | |
| chen12 | lifeless, ok. it is my first bug report to infra... | 09:02 |
| *** humble_ has quit IRC | 09:02 | |
| johnthetubaguy | chen12: if there are missing tags, I would reach out to ttx who will know the best person to ask | 09:03 |
| *** mpavone has quit IRC | 09:03 | |
| lifeless | we have jhesketh on it in -infra | 09:04 |
| chen12 | johnthetubaguy, what is "ttx" | 09:04 |
| jhesketh | lifeless: well looking, no promises | 09:04 |
| ttx | that would be me | 09:05 |
| *** pkonisze has quit IRC | 09:05 | |
| johnthetubaguy | chen12: he is one of the release managers, thiery I always ask him about tags | 09:05 |
| *** pkoniszewski has joined #openstack-nova | 09:05 | |
| ttx | Looks like it's an issue in the GIT farm | 09:05 |
| chen12 | ttx, johnthetubaguy o, good to know that ! Thanks! | 09:05 |
| johnthetubaguy | oh, so I totally got the wrong end of the stuck here, could be a git mirror issue I guess? | 09:06 |
| johnthetubaguy | http://git01.openstack.org:8080/cgit/openstack/nova Missing tag 2014.1.5 & 12.0.0a0 | 09:06 |
| ttx | http://git.openstack.org/cgit/openstack/nova/refs/tags definietly has them | 09:06 |
| ttx | so some stale mirror issue maybe | 09:06 |
| johnthetubaguy | yeah, ignore me sorry, I got overexcited | 09:06 |
| garyk | lifeless: yes, we should move to neutron | 09:06 |
| ttx | git01 is missing them, git02 has them | 09:07 |
| *** aix has joined #openstack-nova | 09:07 | |
| johnthetubaguy | ah, fun fun | 09:07 |
| *** pbrooko has joined #openstack-nova | 09:08 | |
| openstackgerrit | garyk proposed openstack/nova: Cells: add instance cell registration utility to nova-manage https://review.openstack.org/193889 | 09:10 |
| tdurakov | johnthetubaguy: hi John! cvould you look again on spec: https://review.openstack.org/#/c/183057/ I added Alternatives sections | 09:11 |
| openstackgerrit | Stephen Finucane proposed openstack/nova-specs: Add virt-driver CPU thread pinning https://review.openstack.org/194221 | 09:12 |
| openstackgerrit | Stephen Finucane proposed openstack/nova-specs: trivial: Change file permissions for spec https://review.openstack.org/194559 | 09:12 |
| tdurakov | danpb: Hi, seen spec you left in comment to my | 09:14 |
| tonyb | bauzas: ping? | 09:14 |
| tdurakov | danpb: there are no enums too | 09:14 |
| bauzas | tonyb: pong | 09:14 |
| bauzas | tonyb: so I commented https://bugs.launchpad.net/nova/+bug/1456228 | 09:15 |
| openstack | Launchpad bug 1456228 in OpenStack Security Advisory "Trusted vm can be powered on untrusted host" [Undecided,Incomplete] | 09:15 |
| tonyb | bauzas: Yeah moving it out of tree would be nice, perhaps something we could talk about un Tokyo | 09:15 |
| danpb | tdurakov: what spec ? | 09:15 |
| tdurakov | https://review.openstack.org/#/c/183057/5 | 09:15 |
| tdurakov | danpb ^ | 09:15 |
| bauzas | tonyb: as I discussed with johnthetubaguy, I'll open a thread for discussing at least how to change its behaviour | 09:15 |
| tonyb | bauzas: cool. | 09:16 |
| bauzas | tonyb: from a security pov, it sounds weird to rely on something out of Nova for checking if an host is good or not | 09:16 |
| *** park_heijlong has quit IRC | 09:16 | |
| *** bkopilov has quit IRC | 09:17 | |
| bauzas | tonyb: and for example, the bug shows how it's not really useful since the host can be compromised and the instances left in there | 09:17 |
| *** markvoelker has joined #openstack-nova | 09:17 | |
| tonyb | bauzas: Yeah. It's strange. | 09:17 |
| bauzas | tonyb: so I don't want to discuss here on how to cheat with that, but I can see at least one hack vector | 09:17 |
| *** yamahata has joined #openstack-nova | 09:18 | |
| tonyb | bauzas: +1 | 09:18 |
| openstackgerrit | Stephen Finucane proposed openstack/nova-specs: trivial: Change file permissions for spec https://review.openstack.org/194559 | 09:18 |
| openstackgerrit | Stephen Finucane proposed openstack/nova-specs: Add virt-driver CPU thread pinning https://review.openstack.org/194221 | 09:18 |
| bauzas | tonyb: I'm not part of the nova-sec group, do you know how we can safely discuss about that filter and the possible issues ? | 09:19 |
| tonyb | bauzas: so power on != create right and the scheduler is only consiluted on create (and migrate/resize right) | 09:19 |
| bauzas | tonyb: exactly | 09:19 |
| tonyb | bauzas: Cool. | 09:19 |
| bauzas | tonyb: there is one opensource project trying to consider that, btw. | 09:20 |
| bauzas | tonyb: fyi https://wiki.openstack.org/wiki/Watcher | 09:20 |
| tonyb | bauzas: Well there isn't a list right now but an email to me, mikal, alaski, dansmith and johnthetubaguy would be a good starting place | 09:20 |
| bauzas | tonyb: that just means that's out of Nova for checking the current state | 09:21 |
| bauzas | tonyb: ack, will do | 09:21 |
| ekarlso | bauzas: isn't there terracotta too ? or the neat project as it was called | 09:22 |
| *** markvoelker has quit IRC | 09:22 | |
| bauzas | ekarlso: possibly, Congress is also trying to have such of a live check | 09:22 |
| *** afazekas has joined #openstack-nova | 09:22 | |
| tonyb | bauzas: Hmm interesting. | 09:22 |
| ekarlso | bauzas: :| | 09:22 |
| bauzas | ekarlso: my point is just : this is out of Nova | 09:22 |
| *** jcoufal has joined #openstack-nova | 09:23 | |
| *** afazekas has quit IRC | 09:23 | |
| tonyb | bauzas: I was thinking of taking are more "packaging" approach and just move the unmaintained filters into another repo that can be installed alongside nova and the filters will be available. | 09:23 |
| *** pbrooko has quit IRC | 09:23 | |
| bauzas | given an orchestrator called Foo, make sure that Foo has alarms, triggers and actions | 09:23 |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge availability_zone tests between v2 and v2.1 https://review.openstack.org/171017 | 09:23 |
| tonyb | bauzas: not an API for actually talking to the scheduler. | 09:24 |
| *** afazekas has joined #openstack-nova | 09:24 | |
| *** pbrooko has joined #openstack-nova | 09:24 | |
| bauzas | tonyb: that's totally possible, but since TrustedFilter is in-tree, we can't just drop it | 09:24 |
| *** pbrooko has quit IRC | 09:24 | |
| tonyb | bauzas: right. | 09:24 |
| bauzas | tonyb: well, my take on that is that something has to provide stats to the scheduler | 09:25 |
| openstackgerrit | Qin Zhao proposed openstack/nova: Use migration object in compute manager https://review.openstack.org/194565 | 09:25 |
| bauzas | tonyb: for the moment, the "something" is purely nova-compute and nova-api | 09:25 |
| jhesketh | ttx: can you please check if git01 is still missing the tags you were looking at? | 09:25 |
| bauzas | tonyb: but I'm okay if that something becomes "foo" or whatever else | 09:26 |
| bauzas | (call it neutron, cinder or foo) | 09:26 |
| tonyb | bauzas: Yeah. | 09:26 |
| ttx | jhesketh: now shows them at http://git01.openstack.org:8080/cgit/openstack/nova/refs/tags | 09:26 |
| tonyb | bauzas: small steps ;P | 09:26 |
| bauzas | tonyb: eh :) | 09:27 |
| jhesketh | ttx: thanks | 09:27 |
| bauzas | tonyb: the smallest step for Nova I found being claimed is https://www.openstack.org/summit/vancouver-2015/summit-videos/presentation/openstack-is-doomed-and-it-is-your-fault | 09:27 |
| *** pbrooko has joined #openstack-nova | 09:28 | |
| bauzas | tonyb: I assume the next step being to rewrite everything in golang | 09:28 |
| tonyb | bauzas: That's on my list of presentations to watch | 09:28 |
| openstackgerrit | Abhishek Kekane proposed openstack/nova: libvirt: Kill rsync/scp processes before deleting instance https://review.openstack.org/192986 | 09:28 |
| tonyb | bauzas: not rust? ;p | 09:28 |
| bauzas | tonyb: well, just a bit of disclaimer, this presentation is a bit opiniated :p | 09:28 |
| *** bkopilov has joined #openstack-nova | 09:29 | |
| kashyap | So many volume related bugs, ugh. | 09:29 |
| *** Nisha_away has joined #openstack-nova | 09:29 | |
| * kashyap wonders if anyone has any comments on - https://bugs.launchpad.net/nova/+bug/1454252 | 09:30 | |
| openstack | Launchpad bug 1454252 in OpenStack Compute (nova) "Support offline migration of attached volumes when VM is inactive" [High,Confirmed] | 09:30 |
| *** MarcF has quit IRC | 09:30 | |
| tonyb | bauzas: I *never* would have got that from the title :D | 09:30 |
| tonyb | bauzas: Are you coming to the mid-cycle? | 09:30 |
| bauzas | tonyb: unfortunately not :( | 09:30 |
| tonyb | bauzas: Ahh phooey | 09:30 |
| bauzas | tonyb: I was willing to, but my dreams hitted the reality wall | 09:31 |
| bauzas | tonyb: so I'll gently hassle mikal and johnthetubaguy to ask for a remote connectivity :) | 09:31 |
| tonyb | bauzas: Yeah that can happen :( | 09:31 |
| tonyb | bauzas: Sounds good. | 09:31 |
| tonyb | bauzas: I'll poke mikal to bring the microphone that we used last time and try to get a hangout laptop or similar | 09:32 |
| bauzas | tonyb: on my own, I'll set myself on the Rochester TZ | 09:32 |
| tonyb | bauzas: Actually we should ask matt what will be in the room .... | 09:32 |
| *** alexpilotti has joined #openstack-nova | 09:32 | |
| tonyb | bauzas: ekk that'll be hard :( | 09:32 |
| bauzas | tonyb: not really, only like being awake till 1am | 09:33 |
| tonyb | bauzas: hehe okay | 09:33 |
| *** Alexandra_ has joined #openstack-nova | 09:34 | |
| *** Alexandra_ has quit IRC | 09:34 | |
| *** Alexandra_ has joined #openstack-nova | 09:35 | |
| *** yingjun has quit IRC | 09:35 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge security groups tests between v2 and v2.1 https://review.openstack.org/173671 | 09:35 |
| *** jasondotstar has joined #openstack-nova | 09:37 | |
| *** bkopilov has quit IRC | 09:38 | |
| openstackgerrit | Peter Savage proposed openstack/python-novaclient: Added marker functionality to flavours and images https://review.openstack.org/193212 | 09:38 |
| *** bkopilov has joined #openstack-nova | 09:38 | |
| openstackgerrit | Rui Chen proposed openstack/nova: Set task_state=None when booting instance failed https://review.openstack.org/187145 | 09:38 |
| openstackgerrit | Rui Chen proposed openstack/nova: Refactor network API 'get_instance_nw_info' https://review.openstack.org/146036 | 09:38 |
| johnthetubaguy | tonyb: its several rooms joined together, which makes the in room mic situation a bit tricky, from what he mentioned, although one good mic in the middle can go a long way to help I guess | 09:40 |
| *** nihilifer_ has joined #openstack-nova | 09:40 | |
| bauzas | johnthetubaguy: I'm a bit lost with that, does that mean there could be multiple discussions (and agendas) at the same time ? | 09:41 |
| tonyb | johnthetubaguy: Yeah esp if we plan ahead a little so that discussions where remote people like bauzas are in the room with the mic | 09:41 |
| tonyb | anyway /me stops -1'ing patches and checks if he's still married ;P | 09:42 |
| tonyb | bauzas: Thanks for you help. | 09:42 |
| bauzas | tonyb: np, I now have to write down some thougts | 09:43 |
| lxsli | tonyb: pretty sure you are though mikal may be asleep by now | 09:44 |
| rgerganov | claudiub: please remove your -1 on https://review.openstack.org/135625 | 09:44 |
| johnthetubaguy | bauzas: no, its just tricky with AV equipment in that room | 09:44 |
| bauzas | johnthetubaguy: you mean about the room capacity ? | 09:45 |
| *** nihilifer has quit IRC | 09:46 | |
| johnthetubaguy | bauzas: no its just each room has its own mic and speakers, so its tricky when we are spread between the rooms, at least thats my impression | 09:46 |
| *** nihilifer_ has quit IRC | 09:46 | |
| bauzas | johnthetubaguy: aaaah, but all the attendees are in the same room at the same time ?. | 09:46 |
| *** nihilifer has joined #openstack-nova | 09:47 | |
| johnthetubaguy | tonyb: +1 for planning ahead a bit, I was hoping mriedman and mikal would start an etherpad off and things for that soon, I should ping them about that | 09:47 |
| johnthetubaguy | bauzas: yeah, its just the "one" room is actually multiple rooms with some walls opened, I think | 09:47 |
| *** otter768 has joined #openstack-nova | 09:48 | |
| bauzas | johnthetubaguy: oh ack | 09:48 |
| bauzas | johnthetubaguy: anyway, I don't want to slow down the discussions | 09:49 |
| johnthetubaguy | bauzas: I mean it should be OK, well as OK as normal, just not sure we can count on any in room audio stuff, although I am sure we can work around most of that quite easily | 09:49 |
| bauzas | johnthetubaguy: so if I can hear, that's cool, but I should probably only lurk | 09:49 |
| bauzas | johnthetubaguy: since you know, I'm like very slow when speaking | 09:50 |
| danpb | hmm, gerrit seens to have gone awol | 09:50 |
| bauzas | usually, I mean | 09:50 |
| *** nihilifer_ has joined #openstack-nova | 09:50 | |
| johnthetubaguy | danpb: ah, so its not just me, thats some good news | 09:50 |
| kashyap | Ah, I thought it was just me. /me was feverishly refreshing a review like for 4 times. | 09:50 |
| *** oomichi has joined #openstack-nova | 09:51 | |
| danpb | johnthetubaguy: http://downforeveryoneorjustme.com/review.openstack.org says its everyone :-) | 09:51 |
| kashyap | It's back up. It's told me "it's just you" :-) | 09:52 |
| *** deepthi has joined #openstack-nova | 09:53 | |
| *** otter768 has quit IRC | 09:53 | |
| *** xgsa has quit IRC | 09:53 | |
| *** nihilifer has quit IRC | 09:54 | |
| *** oomichi has quit IRC | 09:54 | |
| bauzas | danpb: oooooh, I love that website, so handy | 09:56 |
| *** yassine_ has quit IRC | 09:57 | |
| *** oomichi has joined #openstack-nova | 09:57 | |
| sahid | johnthetubaguy: hello, any chance you take a look at https://review.openstack.org/#/c/139688/ ? | 09:57 |
| sahid | i think we were close | 09:58 |
| johnthetubaguy | sahid: its on my massive queue | 09:58 |
| sahid | ahah... yes :) | 09:58 |
| *** humble__ has quit IRC | 09:58 | |
| johnthetubaguy | sahid: there are currenly 60 specs that need a review | 09:59 |
| sahid | johnthetubaguy: yes i try to help by pointing them close to be merged | 10:00 |
| sahid | already +2ed for example to make a wipe | 10:00 |
| *** chen12 has quit IRC | 10:00 | |
| *** nihilifer_ has quit IRC | 10:00 | |
| *** chen12 has joined #openstack-nova | 10:00 | |
| *** dims has joined #openstack-nova | 10:06 | |
| *** xgsa has joined #openstack-nova | 10:07 | |
| *** ndipanov has joined #openstack-nova | 10:07 | |
| *** oro has joined #openstack-nova | 10:09 | |
| *** IanGovett has joined #openstack-nova | 10:13 | |
| *** fawadkhaliq has quit IRC | 10:15 | |
| *** avishay_ has quit IRC | 10:16 | |
| *** avishay__ has joined #openstack-nova | 10:16 | |
| *** Nisha_away has quit IRC | 10:19 | |
| *** oro has quit IRC | 10:19 | |
| *** thorst has joined #openstack-nova | 10:20 | |
| *** cing has quit IRC | 10:20 | |
| openstackgerrit | Moshe Levi proposed openstack/nova-specs: Add spec to enhance PCI passthrough whitelist to support regex https://review.openstack.org/179577 | 10:25 |
| *** tojuvone has quit IRC | 10:25 | |
| *** varunarya has joined #openstack-nova | 10:26 | |
| *** salv-orlando has quit IRC | 10:29 | |
| johnthetubaguy | sahid: about the set admin root password thing, I think we need to add a note in there about the guest agent, as danpb mentioned in his comments | 10:29 |
| johnthetubaguy | sahid: thanks for the other updates though, I am good to merge it apart from that detail, its different to how xenapi works, so I think its worth calling out (well its a different image property, I think, because its a different required guest agent) | 10:30 |
| *** oomichi has quit IRC | 10:31 | |
| sudipto | johnthetubaguy, do you consider this good for merge now: https://review.openstack.org/#/c/180983/ ? | 10:31 |
| sudipto | johnthetubaguy, sorry if it's already in your queue :) | 10:31 |
| johnthetubaguy | sudipto: its on my list | 10:31 |
| sudipto | johnthetubaguy, thx. | 10:31 |
| markus_z | kashyap: FYI: http://docs-draft.openstack.org/21/191021/4/check/gate-nova-docs/edb3901//doc/build/html/vmstates.html | 10:38 |
| garyk | danpb: would it be possible that you please look at https://review.openstack.org/192675 | 10:38 |
| kashyap | markus_z: Yep, thanks. | 10:38 |
| *** tobe has quit IRC | 10:39 | |
| *** ociuhandu has joined #openstack-nova | 10:40 | |
| *** doron_afk has quit IRC | 10:40 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: WIP: Deprecate TrusterFilter https://review.openstack.org/194592 | 10:41 |
| bauzas | johnthetubaguy: tonyb: FYI http://lists.openstack.org/pipermail/openstack-dev/2015-June/067766.html | 10:42 |
| * bauzas going out to lunch | 10:42 | |
| * johnthetubaguy is now thinking about lunch... mmm, food, yummy | 10:43 | |
| sahid | danpb: about hw_qemu_guest_agent | 10:48 |
| * StevenK is currently cooking spaghetti for dinner if that helps johnthetubaguy at all ... | 10:48 | |
| *** pbrooko has quit IRC | 10:49 | |
| openstackgerrit | Mark Goddard proposed openstack/nova: Refresh instance info cache within lock https://review.openstack.org/194597 | 10:49 |
| sahid | 'im not sure it is relevant to ask user to set this property, why not just ccatach exception raised from libvirt if the agent in not running? | 10:49 |
| johnthetubaguy | sahid: so from what danpb said, the user has to do that, else libvirt will just not contact the guest agent, or something like that | 10:50 |
| johnthetubaguy | sahid: seems to be what happens already | 10:51 |
| *** doron_afk has joined #openstack-nova | 10:51 | |
| johnthetubaguy | sahid: it should really be something the deployer sets on the base images and gets inherited by snapshots, so it is not "too" exposed to the user, in theory, although ideally we could just detect all this stuff | 10:51 |
| johnthetubaguy | StevenK: :) | 10:52 |
| sahid | johnthetubaguy: hum... we do a manual check in the driver before to invoke libvirt | 10:52 |
| openstackgerrit | Mark Goddard proposed openstack/nova: Refresh instance info cache within lock https://review.openstack.org/194597 | 10:53 |
| sahid | yes i guess that makes sense we can do the job in the API to return the error directly | 10:53 |
| sahid | if image does not have the requirement | 10:53 |
| johnthetubaguy | sahid: so I think thats what XenAPI does in some cases, but I don't 100% remember | 10:54 |
| johnthetubaguy | sahid: can't be done in the API though, its driver specific | 10:54 |
| *** coolsvap is now known as coolsvap|away | 10:55 | |
| sahid | johnthetubaguy: yes we need some specific information like virt_type which is driver specific | 10:55 |
| *** yamahata has quit IRC | 10:55 | |
| johnthetubaguy | sahid: so this sucks, its a cast, so the API just silently returns success: https://github.com/openstack/nova/blob/master/nova/virt/xenapi/vmops.py#L1309 | 10:55 |
| johnthetubaguy | sahid: my bad, it is already kinda handled: https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L2996 | 10:56 |
| *** lyarwood has quit IRC | 10:57 | |
| johnthetubaguy | sahid: I would just follow the existing pattern for now | 10:57 |
| *** lyarwood has joined #openstack-nova | 10:57 | |
| sahid | johnthetubaguy: ok no problem let's do that | 10:57 |
| *** gszasz has joined #openstack-nova | 10:58 | |
| openstackgerrit | sahid proposed openstack/nova-specs: libvirt: set admin root password https://review.openstack.org/187509 | 10:59 |
| *** jamielennox is now known as jamielennox|away | 11:00 | |
| johnthetubaguy | sahid: do ping me when you have that updated, and I can ack that to speed things up | 11:03 |
| *** markvoelker has joined #openstack-nova | 11:06 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Add rootwrap daemon mode support https://review.openstack.org/180695 | 11:07 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add timing information for rootwrap daemon https://review.openstack.org/194602 | 11:07 |
| sahid | johnthetubaguy: it's done | 11:07 |
| sahid | in the user impact section | 11:07 |
| *** hanzhang has quit IRC | 11:07 | |
| johnthetubaguy | sahid: you got the link handy? | 11:07 |
| sahid | https://review.openstack.org/#/c/187509/6/specs/liberty/approved/libvirt-set-admin-password.rst,cm | 11:07 |
| *** liusheng has quit IRC | 11:08 | |
| *** RuiChen has quit IRC | 11:08 | |
| *** liusheng has joined #openstack-nova | 11:08 | |
| *** kiwik-chenrui has joined #openstack-nova | 11:08 | |
| *** hanzhang has joined #openstack-nova | 11:08 | |
| johnthetubaguy | sahid: I was thinking a bit more detail, but I think that will become very clear when you start coding, so I will go for that one, thanks | 11:09 |
| sahid | oops :) | 11:10 |
| *** tojuvone has joined #openstack-nova | 11:10 | |
| *** markvoelker has quit IRC | 11:11 | |
| *** Alexandra_ has quit IRC | 11:11 | |
| *** mpavone has joined #openstack-nova | 11:11 | |
| *** irenab has quit IRC | 11:12 | |
| *** irenab has joined #openstack-nova | 11:13 | |
| *** ZZelle has quit IRC | 11:14 | |
| *** ZZelle has joined #openstack-nova | 11:14 | |
| *** varunarya has quit IRC | 11:15 | |
| openstackgerrit | Daniel Berrange proposed openstack/nova-specs: VIF port config versioned objects and driver plugin library https://review.openstack.org/193668 | 11:15 |
| *** aix has quit IRC | 11:15 | |
| prashup | sahid | 11:17 |
| prashup | i was making the change in v2.1 | 11:17 |
| prashup | but i dont know how to test v2.1 from CLI | 11:17 |
| prashup | help !!!! | 11:17 |
| prashup | @sahid ^^^ | 11:17 |
| sahid | prashup: ? | 11:19 |
| abhishekk | prashup: you need to use curl | 11:20 |
| johnthetubaguy | prashup: I think there is a spec talk talks about adding the CLI support for v2.1, not sure its approved yet | 11:20 |
| prashup | there is no way to test v2.1 from CLI ?? | 11:20 |
| abhishekk | prashup: no as of now | 11:20 |
| prashup | ok | 11:21 |
| prashup | i m new to nova, openstack and this would be mine first checkin. | 11:21 |
| prashup | can you guys tell me how to test it through curl request? | 11:21 |
| abhishekk | prashup: ohhh | 11:21 |
| prashup | or any doc i can go through | 11:21 |
| prashup | ? | 11:21 |
| abhishekk | prashup: api.openstack.org might be useful | 11:22 |
| openstackgerrit | Daniel Berrange proposed openstack/nova-specs: VIF port config versioned objects and driver plugin library https://review.openstack.org/193668 | 11:22 |
| prashup | thanks | 11:22 |
| abhishekk | prashup: which command you want to run? | 11:22 |
| prashup | nova boot | 11:22 |
| prashup | boot from volume to be specific | 11:23 |
| *** baoli has joined #openstack-nova | 11:23 | |
| abhishekk | prashup: ok, I will give you the curl command, you need to change token and host in that command | 11:23 |
| *** yassine_ has joined #openstack-nova | 11:24 | |
| prashup | how would i get token ? when i see nova CLI using debug flag i see it takes care of it | 11:24 |
| * johnthetubaguy runs away to lunch | 11:24 | |
| *** oro has joined #openstack-nova | 11:24 | |
| abhishekk | prashup: keystone token-get, when you are running curl command then you need to generate token by yourself and pass it to curl | 11:27 |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Adds Hyper-V Shielded VMs spec https://review.openstack.org/192892 | 11:28 |
| *** aix has joined #openstack-nova | 11:28 | |
| *** jcoufal has quit IRC | 11:29 | |
| *** salv-orlando has joined #openstack-nova | 11:30 | |
| abhishekk | prashup: curl -g -i -X POST http://10.69.4.177:8774/v2.1/cdce26c324134f65aa2424f25574f1a5/os-volumes_boot -H "User-Agent: python-novaclient" -H "Content-Type: application/json" -H "Accept: application/json" -H "X-Auth-Token: 0a9563261c04449bbd63ecc9af2d873d" -d '{"server": {"name": "nova1", "imageRef": "", "block_device_mapping_v2": [{"source_type": "volume", "delete_on_termination": false, "boot_index": 0, "uuid": "2edda400-6b10-4118-bb32-1d16 | 11:30 |
| abhishekk | d5c5c1a0", "destination_type": "volume"}], "flavorRef": "2", "max_count": 1, "min_count": 1}}' | 11:30 |
| abhishekk | prashup: replace 10.69.4.177 with your controller node ip and "0a9563261c04449bbd63ecc9af2d873d" with id of "keystone token-get" output | 11:31 |
| *** lyarwood has quit IRC | 11:33 | |
| *** lyarwood has joined #openstack-nova | 11:33 | |
| *** salv-orlando has quit IRC | 11:36 | |
| prashup | sure thanks abhishekk | 11:37 |
| abhishekk | prashup: also you need to pass voume_id in uuid :) | 11:37 |
| prashup | yeah that i will | 11:38 |
| prashup | :) | 11:38 |
| abhishekk | prashup: great | 11:38 |
| prashup | i will also need to add size | 11:38 |
| prashup | ;) | 11:38 |
| *** kiran-r has joined #openstack-nova | 11:38 | |
| *** ociuhandu has quit IRC | 11:39 | |
| *** doron_afk is now known as doron_ | 11:39 | |
| abhishekk | :) | 11:41 |
| *** lucasagomes is now known as lucas-hungry | 11:42 | |
| *** pbrooko has joined #openstack-nova | 11:42 | |
| *** deepthi has quit IRC | 11:43 | |
| *** zhenguo has quit IRC | 11:43 | |
| *** novaUser has joined #openstack-nova | 11:44 | |
| *** novaUser has left #openstack-nova | 11:44 | |
| *** sdake has joined #openstack-nova | 11:46 | |
| *** prashup has quit IRC | 11:46 | |
| *** sdake_ has joined #openstack-nova | 11:47 | |
| *** otter768 has joined #openstack-nova | 11:49 | |
| *** dprince has joined #openstack-nova | 11:49 | |
| *** otter768 has quit IRC | 11:50 | |
| *** otter768 has joined #openstack-nova | 11:50 | |
| *** sdake has quit IRC | 11:50 | |
| *** avishay__ has quit IRC | 11:53 | |
| *** salv-orlando has joined #openstack-nova | 11:54 | |
| *** sdake_ has quit IRC | 11:54 | |
| *** husanu has quit IRC | 11:56 | |
| *** fawadkhaliq has joined #openstack-nova | 11:57 | |
| *** husanu has joined #openstack-nova | 11:58 | |
| *** ociuhandu has joined #openstack-nova | 11:59 | |
| *** mpavone has quit IRC | 12:00 | |
| *** otter768 has quit IRC | 12:00 | |
| openstackgerrit | Petrut Lucian proposed openstack/nova: Hyper-V: Fixes method retrieving free SCSI controller slot on V1 https://review.openstack.org/194616 | 12:02 |
| *** moshele has quit IRC | 12:02 | |
| *** pask81 has joined #openstack-nova | 12:05 | |
| *** markvoelker has joined #openstack-nova | 12:05 | |
| *** mtanino has joined #openstack-nova | 12:05 | |
| *** jasondotstar has quit IRC | 12:06 | |
| *** kes has joined #openstack-nova | 12:06 | |
| openstackgerrit | Peter Savage proposed openstack/python-novaclient: Added marker functionality to flavours and images https://review.openstack.org/193212 | 12:06 |
| *** sudipto has quit IRC | 12:07 | |
| *** romainh has quit IRC | 12:09 | |
| markus_z | anybody here from XEN? | 12:10 |
| *** obondarev has quit IRC | 12:10 | |
| openstackgerrit | sean mooney proposed openstack/nova-specs: Filter - aggregate flavour extra spec affinity filter https://review.openstack.org/183876 | 12:11 |
| BobBall | sure markus_z | 12:12 |
| openstackgerrit | Alexandre Levine proposed openstack/nova-specs: delete_on_termination modification spec https://review.openstack.org/194144 | 12:12 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Added server tags support in nova-api https://review.openstack.org/128940 | 12:13 |
| markus_z | BobBall: Cool, can you maybe help me to understand the "host maintenance mode"? | 12:13 |
| BobBall | Best to ask that in #xen-api | 12:13 |
| markus_z | BobBall: Could it be that the "evacuate instances from host" is the same: http://docs.openstack.org/developer/nova/support-matrix.html#operation_evacuate | 12:13 |
| BobBall | oh | 12:13 |
| BobBall | Not really | 12:13 |
| BobBall | Well - do you mean host maintenance mode in XAPI? | 12:14 |
| BobBall | or in openstack in general? | 12:14 |
| *** pask81 has quit IRC | 12:14 | |
| samueldmq | sdague: morning | 12:14 |
| *** moshele has joined #openstack-nova | 12:15 | |
| sdague | samueldmq: morning | 12:15 |
| markus_z | BobBall: I'm confused what the feature means in the support matrix from OpenStack. | 12:15 |
| samueldmq | sdague: I'd like to agree with you the points in the meeting today that would be the most productive for both nova and keystone sides | 12:15 |
| markus_z | The title and the text doesn't fit together | 12:15 |
| *** raildo has joined #openstack-nova | 12:15 | |
| *** claudiub has quit IRC | 12:15 | |
| sdague | samueldmq: sounds good, what are you thinking? | 12:15 |
| samueldmq | sdague: I think i) unified vs not unified, and then if not unified: CMS or /policy ? | 12:16 |
| BobBall | Got it. The feature is effectively that if you want to do maintenance on a host then you can evacuate all instances and mark the host as being in maintenance mode. Once all instances are off you can then power down the host and replace a HDD or whatever then bring it out of this "maintenance mode" | 12:16 |
| markus_z | BobBall: The funny thing is, the XEN driver does an evacuate when setting the host in maintenance mode: https://github.com/openstack/nova/blob/master/nova/virt/xenapi/host.py#L56 | 12:16 |
| sdague | samueldmq: define what you mean by unified or not unified? | 12:16 |
| samueldmq | sdague: (I am not assuming we've agreed on anything, just exploring the possibilities) | 12:16 |
| samueldmq | sdague: unified policy, what ayoung wants | 12:16 |
| markus_z | BobBall: OK, that means the title in the support matrix is wrong, isn't it? | 12:17 |
| *** ctrath has joined #openstack-nova | 12:17 | |
| sdague | unified meaning that keystone owns all the policy files? | 12:17 |
| BobBall | I think the official name is "Evacuate host" rather than putting it into maintenance mode. johnthetubaguy? | 12:17 |
| samueldmq | sdague: yes somehow, and has all those problems with microversions, people running on master, etc | 12:17 |
| johnthetubaguy | BobBall: we have things in the API, I believe | 12:17 |
| sdague | samueldmq: right, which I think it is a non starter :) | 12:17 |
| johnthetubaguy | BobBall: I mean, we have both things in the API | 12:17 |
| sdague | but sure, we can discuss that part first | 12:18 |
| samueldmq | sdague: he isn't still convinced, so the meeting is a good place to talk all together, and have an agreement | 12:18 |
| samueldmq | sdague: yes but I think this is the major disagreement we have right now | 12:18 |
| johnthetubaguy | markus_z: I want to remove that xenapi stuff really, its totally untested, and just doesn't map very well | 12:18 |
| *** VW_ has joined #openstack-nova | 12:18 | |
| markus_z | BobBall: I tried to clarify it via ML http://openstack.markmail.org/thread/nd37hdnop3zelzbk | 12:18 |
| *** mgarza_ has joined #openstack-nova | 12:19 | |
| BobBall | Well my interpretation of that feature is that the header is correct | 12:19 |
| samueldmq | sdague: besides that, we should be ok in something like we've in those sequence diagrams (not assuming /policy or CMS) | 12:19 |
| samueldmq | sdague: sounds good this first point ? | 12:19 |
| sdague | yeh, that sounds reasonable. Honestly though, I feel like it doesn't turn into an execution plan. | 12:19 |
| sdague | because an execution plan is "what's the first incremental step forward" | 12:20 |
| BobBall | Ah - joy. I see that this is another case of something being implemented in two places historically :) | 12:20 |
| johnthetubaguy | BobBall: evacuate is nothing to do with maintenance mode stuff | 12:20 |
| BobBall | No? | 12:20 |
| BobBall | OK then I'm confused too | 12:20 |
| markus_z | johnthetubaguy: I still don't get my head around what the expectations are when starting "host maintenance mode". | 12:20 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: libvirt: mark NUMA huge page mappings as shared access https://review.openstack.org/176360 | 12:20 |
| samueldmq | sdague: no that's not the execution plan at all, that's the agreement on the roadmap | 12:21 |
| johnthetubaguy | markus_z: its a complete mess around host maintenance mode, frankly the current feature is totally usable, IMHO | 12:21 |
| *** oro has quit IRC | 12:21 | |
| johnthetubaguy | BobBall: evacuate is about when a host is dead, and to do your best to start the instance on a new host | 12:21 |
| sdague | samueldmq: so... I feel like it is both too detailed and not detailed enough. I think the roadmap shouldn't be sequence diagrams | 12:22 |
| johnthetubaguy | BobBall: if its not boot from volume, it will look to the user like a rebuild (i.e. they loose all their data) | 12:22 |
| sdague | the roadmap should be a vision | 12:22 |
| BobBall | OK - so there is no OpenStack way of saying "Let's take this host out of usage for some maintenance" | 12:22 |
| BobBall | (apart from this maintenance mode for XenAPI) | 12:22 |
| sdague | and then we think about the things that could be done to get us closer to that vision, that would have incremental benefits, and support getting closer to it over time | 12:22 |
| samueldmq | sdague: ok the sequence diagrams tehre are the scope for L | 12:23 |
| johnthetubaguy | sdague: samueldmq: I am very curious about the context around roadmaps? | 12:23 |
| samueldmq | sdague: what I am trying to do there is to i) roadmap and ii) scope for L, based on roadmap | 12:23 |
| samueldmq | sdague: if that makes snese | 12:23 |
| * bauzas notices review.o.o is still on holiday | 12:23 | |
| samueldmq | johnthetubaguy: hi, we're talking about dynamic policies corss-porject effort :) | 12:23 |
| BobBall | bauzas: WFM | 12:24 |
| *** claudiub has joined #openstack-nova | 12:24 | |
| *** kiran-r has quit IRC | 12:24 | |
| johnthetubaguy | samueldmq: ah, gotcha | 12:24 |
| sdague | samueldmq: sure, I guess, we disagree on the fact that those sequence diagrams are scope for L | 12:24 |
| bauzas | BobBall: fixed just now | 12:24 |
| samueldmq | johnthetubaguy: we've some initial iteration on ideas from keystone side (https://wiki.openstack.org/wiki/DynamicPolicies) | 12:24 |
| bauzas | BobBall: was not working like 5 secs before | 12:24 |
| BobBall | heh :) | 12:24 |
| samueldmq | johnthetubaguy: and now we come here to hear you guys, your needs, requirements and ideas | 12:24 |
| johnthetubaguy | bauzas: its been really flakey this morning | 12:24 |
| bauzas | johnthetubaguy: just joined -infra, saw some discussion in there, will see | 12:25 |
| samueldmq | sdague: that's an initial iteration, that will change once we introduce nova requirements/needs/ideas | 12:25 |
| samueldmq | sdague: don't see that as something wrote in stone, we're open, remember | 12:25 |
| *** zzz_pczesno is now known as pczesno | 12:25 | |
| sdague | samueldmq: ok, sure | 12:25 |
| *** ivasev has joined #openstack-nova | 12:25 | |
| samueldmq | sdague: so point i) we get agreement on wheter unify or not | 12:25 |
| *** pask81 has joined #openstack-nova | 12:25 | |
| johnthetubaguy | samueldmq: so I am actually very against policy being updated via API, I want it to happen in sync with my code deploy really, and like all the other configuration I push out | 12:26 |
| johnthetubaguy | samueldmq: now a lot of the other parts are super useful | 12:26 |
| *** mpavone has joined #openstack-nova | 12:26 | |
| *** aysyanne has joined #openstack-nova | 12:26 | |
| *** agentle has joined #openstack-nova | 12:26 | |
| *** pask81 has quit IRC | 12:26 | |
| *** pask81 has joined #openstack-nova | 12:27 | |
| johnthetubaguy | sdague: we should totally tidy up our policy plan, and share that with keystone, seems like our ideas our converging | 12:27 |
| markus_z | johnthetubaguy: BobBall: Should the title "Evacuate instances from host" be "Set host into maintenance mode"? This would mean a "missing" for s390x too. | 12:27 |
| sdague | johnthetubaguy: well, it's all colliding here | 12:27 |
| sdague | the keystone meeting today is going to be part of this discussion | 12:27 |
| markus_z | I put a "complete" into it because I thought the "nova evacuate" calls are meant. | 12:27 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: add os_admin_user to use with set admin password https://review.openstack.org/185921 | 12:27 |
| openstackgerrit | sahid proposed openstack/nova: libvirt: enchance libvirt to set admin password https://review.openstack.org/185910 | 12:27 |
| johnthetubaguy | samueldmq: my feedback on that wiki is mostly that its not clear what problems you are trying to solve, but maybe thats just how I think about things, not sure | 12:28 |
| johnthetubaguy | sdague: ah, OK | 12:28 |
| samueldmq | sdague: point ii) we try to redefine scope for L | 12:28 |
| *** edmondsw has joined #openstack-nova | 12:28 | |
| BobBall | markus_z: That'd work for me - clearly a quicker fix than working out what to do properly with the XenAPI implementation. | 12:28 |
| johnthetubaguy | markus_z: BobBall: you are mixing up two different things here I think, I would talk about "nova evacuate <server-uuid>" and its per instance not per host | 12:29 |
| bauzas | johnthetubaguy: that's sometimes confusing because the CLI provides an host-evacuate stuff which calls sequentially the evacuate API | 12:30 |
| samueldmq | johnthetubaguy: I don't understand what 'very against policy being updated via API', so the policy should not be able to be updated via API ? | 12:30 |
| markus_z | johnthetubaguy: There is "nova host-evacuate <host>" too | 12:30 |
| johnthetubaguy | markus_z: BobBall: there are plans to deprecating that current hypervisor metrics to be replaced by a new more explicit "feature classification" system, but I haven't started work on that yet | 12:30 |
| johnthetubaguy | bauzas: agreed | 12:30 |
| *** atuvenie has joined #openstack-nova | 12:30 | |
| samueldmq | sdague: is that something we're still missing in the discussion ^? | 12:30 |
| johnthetubaguy | samueldmq: I think its important to update policy at the same time you update your code, and A/B test policy, etc, like its possible today | 12:31 |
| sdague | samueldmq: I think it comes down to the fact there is no vision here | 12:31 |
| *** aix has quit IRC | 12:31 | |
| johnthetubaguy | samueldmq: now maybe some of that is silly, but its quite a change that I suspect deployers will have a strong opinion on, I guess | 12:32 |
| *** fawadkhaliq has quit IRC | 12:32 | |
| *** romainh has joined #openstack-nova | 12:32 | |
| sdague | and policy means different things to different parts of the openstack stack. For instance, in nova we're mostly thinking about policy in terms of very granular function points. And all my conversations in keystone talk about the benefits here in expanding the custom roles that can exist. | 12:32 |
| sdague | johnthetubaguy: so, if this was in layers | 12:33 |
| johnthetubaguy | sdague: good point, I like the idea of keystone "owning" the definition of a role, FWIW | 12:33 |
| sdague | policy in code, policy in patch files, policy applied dynamically via API | 12:33 |
| sdague | you wouldn't be opposed to an API right? | 12:33 |
| sdague | you are just opposed to an API as base source of truth | 12:33 |
| johnthetubaguy | sdague: yeah, thats a good way of putting it | 12:34 |
| johnthetubaguy | sdague: dymanic overide seems handy, but seems like a massive PITA for most of the changes | 12:34 |
| *** ivasev has quit IRC | 12:34 | |
| johnthetubaguy | sdague: mostly wearing my large deployer hat for that bit of the conversation | 12:35 |
| samueldmq | sdague: nice, the way johnthetubaguy I understood him liek against the API to changing policy rules/roles dynamically, and that basically would be against all dynamic policy work | 12:35 |
| *** agentle has quit IRC | 12:35 | |
| *** agentle has joined #openstack-nova | 12:36 | |
| johnthetubaguy | samueldmq: yeah, I am cool with the dynamic, just moving to all dynamic seems like a disaster | 12:36 |
| *** bkopilov has quit IRC | 12:36 | |
| BobBall | johnthetubaguy: Don't suppose you know what step is in progress with spawn stalled at 70%? | 12:36 |
| johnthetubaguy | (but I don't feel like I am not seeing the full picture here, yet) | 12:36 |
| *** ajayaa has quit IRC | 12:37 | |
| *** MVenesio has joined #openstack-nova | 12:37 | |
| samueldmq | johnthetubaguy: we're basically moving from the old .json files, and allowing admins to customize via API, getting warnings if changing the defaults (cc sdague) | 12:37 |
| johnthetubaguy | BobBall: agent is 80% I guess its the one before | 12:37 |
| sdague | johnthetubaguy: yeh, I think the issue is there really isn't a full picture as of yet | 12:37 |
| samueldmq | johnthetubaguy: also, we could do some validation on the rules, for example, roles used in policy rules MUST exist | 12:37 |
| sdague | so I picked together all the things I thought were must haves for people, and built my own picture, and am just telling that to everyone :) | 12:37 |
| johnthetubaguy | sdague: yep, I have something similar in my head | 12:38 |
| samueldmq | johnthetubaguy: basically we would be guiding people to write policies the right way | 12:38 |
| johnthetubaguy | sdague: although, to you point, its a different partial picture | 12:38 |
| samueldmq | johnthetubaguy: also, better defaults policies in the oopenstack ecosystem is on the roadmap :) | 12:38 |
| *** tojuvone has quit IRC | 12:38 | |
| *** tojuvone has joined #openstack-nova | 12:38 | |
| markus_z | johnthetubaguy: I'm very interested in that new feature classification system. If you need support, call me. | 12:39 |
| *** kylek3h has quit IRC | 12:39 | |
| johnthetubaguy | samueldmq: so I don't see the wiki having a listing of problems we are trying to solve, I mean it sounds cool, but I am worried about it | 12:39 |
| johnthetubaguy | markus_z: thanks for the offer, I want to get that kicked off, but yes, will need lots of help with that :) | 12:39 |
| kashyap | markus_z: Where is the feature classifcation stuff? | 12:39 |
| samueldmq | johnthetubaguy: ok so the issues behind the usecases, right ? | 12:39 |
| *** MVenesio has quit IRC | 12:39 | |
| johnthetubaguy | kashyap: doesn't exist yet, its in my head, as discussed at the summit | 12:39 |
| kashyap | Ah, okay. | 12:40 |
| johnthetubaguy | samueldmq: not really, ignore the use cases, talk about issues deployers and devs have today, then see where that goes | 12:40 |
| markus_z | Is there an etherpad of the discussion? I couldn't attent the last summit. I'll be at the midcycle event in July. | 12:40 |
| markus_z | johnthetubaguy: ^ | 12:41 |
| johnthetubaguy | markus_z: kashyap: https://wiki.openstack.org/wiki/Design_Summit/Liberty/Etherpads#Nova then see: https://etherpad.openstack.org/p/YVR-nova-functional-testing-feature-classification | 12:41 |
| markus_z | thanks | 12:41 |
| openstackgerrit | Alexandre Levine proposed openstack/nova-specs: delete_on_termination modification spec https://review.openstack.org/194144 | 12:41 |
| *** prashup has joined #openstack-nova | 12:42 | |
| *** dave-mccowan has joined #openstack-nova | 12:42 | |
| *** boris-42 has quit IRC | 12:42 | |
| johnthetubaguy | samueldmq: the problem is we really need to capture how our users use the current system, deploying policy just like any other configuration, different policy for different API endpoints, and all that kind of thing | 12:42 |
| *** lucas-hungry is now known as lucasagomes | 12:43 | |
| *** mwagner_afk has joined #openstack-nova | 12:43 | |
| johnthetubaguy | samueldmq: certainly we have four types of nova-api nodes, two types of keystone, with different policy for each, three or four types of compute "admin" / "support", etc, etc | 12:43 |
| *** ivasev has joined #openstack-nova | 12:43 | |
| johnthetubaguy | samueldmq: I mean there is an argument that says most of that is totally nuts, but its what people are doing today | 12:43 |
| samueldmq | johnthetubaguy: oh sure, different endpoints can have different policies | 12:44 |
| johnthetubaguy | samueldmq: and when you deploy the new code, you have old APIs with the old policy, and new APIs with the new policy, and you drain and go from one to the other, each having been though QA independently to check they work, etc, etc. | 12:45 |
| johnthetubaguy | all part of the same endpoint, but behind a loadbalancer that deals with the draining, etc | 12:45 |
| samueldmq | johnthetubaguy: if they're behind a loadbalancer, they will have the same endpoint URL, which implies they'll all have the same policy | 12:46 |
| samueldmq | johnthetubaguy: we'll take care of this fetching/caching strategy at keystonemiddleware | 12:46 |
| johnthetubaguy | samueldmq: they don't have the same policy during the deployment of new code | 12:46 |
| *** aix has joined #openstack-nova | 12:47 | |
| samueldmq | johnthetubaguy: how can you provide different versions of the same service behind a, lets's say, HAProxy ? | 12:47 |
| johnthetubaguy | samueldmq: older code runs on older (tested) policy, new code starts using the new policy, (think live upgrades). | 12:47 |
| *** rfolco has joined #openstack-nova | 12:47 | |
| johnthetubaguy | samueldmq: the REST API is backwards compatible, so you can slowly update the nodes without people noticing | 12:48 |
| samueldmq | johnthetubaguy: the policy should be backwards compatible as well, shoudln't it ? | 12:49 |
| johnthetubaguy | samueldmq: yes, but its not QAed like that currently, as it goes out with the new code, which is nice and simple, just like how all other config works | 12:49 |
| *** pask81 has quit IRC | 12:50 | |
| samueldmq | johnthetubaguy: that will be very easy to do with microversions, and how we represent them in the polciy file | 12:50 |
| johnthetubaguy | samueldmq: not saying it has to be like that always, just saying, you are fundamentally changing how our deployment process would have to work, so thats going to take time to evolve | 12:50 |
| *** pask81 has joined #openstack-nova | 12:50 | |
| samueldmq | johnthetubaguy: let's say we do like : 'create_server': {'2.0':'role:admin', '2.1':'role:admin or role:server_admin'} | 12:50 |
| samueldmq | johnthetubaguy: we shouldn't change the policy for a given microvesion | 12:51 |
| *** amotoki has joined #openstack-nova | 12:51 | |
| *** coolsvap|away is now known as coolsvap | 12:51 | |
| samueldmq | johnthetubaguy: to keep compatibility, it is shipped with the api microversion altogether | 12:51 |
| samueldmq | johnthetubaguy: if that makes sense | 12:51 |
| johnthetubaguy | samueldmq: its more about the QA process, you test the new policy, you A/B test it on your production nodes, then finally flip it across | 12:51 |
| *** salv-orlando has quit IRC | 12:52 | |
| johnthetubaguy | samueldmq: that can be independent or directly tied to code changes | 12:52 |
| johnthetubaguy | now its a use case we might want to exclude, but we need a transition period, etc, etc | 12:52 |
| johnthetubaguy | samueldmq: all as I am saying is having an API to change policy isn't a requirement I have as a deployer, it would be much more hard work than just a .json file that my ansible scripts update in a nice controlled way | 12:53 |
| *** atuvenie_ has joined #openstack-nova | 12:54 | |
| johnthetubaguy | anyways, I should really find a way of writing this up, IRC is getting no where here | 12:54 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: libvirt: set caps on maximum live migration time https://review.openstack.org/162254 | 12:55 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: libvirt: support management of downtime during migration https://review.openstack.org/162253 | 12:55 |
| *** toure has joined #openstack-nova | 12:56 | |
| samueldmq | johnthetubaguy: yes please do, keystone is very open to hear to potential corner cases, concerns in this subject | 12:57 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Add rootwrap daemon mode support https://review.openstack.org/180695 | 12:57 |
| samueldmq | johnthetubaguy: that's what sdague did, he has some concerns/ideas, and we'll be discussing some of them today | 12:57 |
| openstackgerrit | Alex Xu proposed openstack/nova-specs: Allow specifying cpu model/features in flavor extra-specs or image metadata https://review.openstack.org/168982 | 12:57 |
| *** atuvenie has quit IRC | 12:57 | |
| openstackgerrit | Daniel Berrange proposed openstack/nova: ironic: convert driver to use nova.objects.ImageMeta https://review.openstack.org/189362 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: objects: convert config drive to use ImageMeta object https://review.openstack.org/189323 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: libvirt: convert driver to use nova.objects.ImageMeta https://review.openstack.org/188119 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: libvirt: make instance compulsory in blockinfo APIs https://review.openstack.org/188118 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: utils: ignore block device mapping in system metadata https://review.openstack.org/192627 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: hyperv: convert driver to use nova.objects.ImageMeta https://review.openstack.org/189302 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: objects: add virtual 'image_meta' property to Instance object https://review.openstack.org/190145 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: virt: convert hardware module to use nova.objects.ImageMeta https://review.openstack.org/188789 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: libvirt: convert to use instance.image_meta property https://review.openstack.org/190146 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: compute: convert manager to use nova.objects.ImageMeta https://review.openstack.org/189685 | 12:58 |
| openstackgerrit | Daniel Berrange proposed openstack/nova: xenapi: convert driver to use nova.objects.ImageMeta https://review.openstack.org/188790 | 12:58 |
| samueldmq | johnthetubaguy: the idea is to have a big win in this dynamic policy subject all togehter, if you have important concerns, please come with them from the beggining | 12:58 |
| samueldmq | johnthetubaguy: you're also invited to the keystone meeting today (18 utc) | 12:58 |
| johnthetubaguy | samueldmq: I guess my thing, is I hope we don't have to loose static policy, for where thats useful | 12:58 |
| samueldmq | johnthetubaguy: so you still want ot have the old policy.json files ? and always be customizing inside them using an out-of-band mechanism? | 12:59 |
| samueldmq | johnthetubaguy: no change to what we have today in the way they're distributed/customized? | 12:59 |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova-specs: libvirt: virtio-net multiqueue https://review.openstack.org/128825 | 13:00 |
| johnthetubaguy | samueldmq: the idea in my head seems to want the old policy files, basically, but with the addition of dynamic policy over-rides | 13:00 |
| sdague | samueldmq: right, so that's the part which isn't all that clear, and why we need a big vision of where we are going. | 13:00 |
| sdague | and a concept level description | 13:01 |
| johnthetubaguy | samueldmq: over time, you could consider removing the static policy files, if people don't want them anymore, so you get a smooth transition from the old world to the new world | 13:01 |
| *** coolsvap is now known as coolsvap|away | 13:01 | |
| samueldmq | johnthetubaguy: sdague so keep the static policy files consistent with the defaults in the code in nova | 13:01 |
| samueldmq | johnthetubaguy: so if people don't want to use policy.json they enable dynamic policies | 13:01 |
| samueldmq | we plan to be backwards compatible | 13:02 |
| *** moshele has quit IRC | 13:02 | |
| *** xyang1 has joined #openstack-nova | 13:02 | |
| *** mpavone has quit IRC | 13:02 | |
| johnthetubaguy | samueldmq: I want the default policy file empty, well full of comments, generated from the policy thats in the code, but you want to be able to discover the policy and over-ride in the API, ideally | 13:02 |
| samueldmq | johnthetubaguy: so policies would still be kept as today, nothing would be dynamic besides the generation of the policy.json in the nova side | 13:03 |
| samueldmq | sdague: yes I think we need that description | 13:03 |
| samueldmq | johnthetubaguy: I'll write up a second version of that api, starting from the problems we're trying to solve | 13:04 |
| johnthetubaguy | samueldmq: I like the idea of over-riding the policy via an admin API call, but that would over-ride the existing systems | 13:04 |
| johnthetubaguy | samueldmq: that would be great | 13:04 |
| samueldmq | johnthetubaguy: sure that way we will be syncrhonizing | 13:04 |
| samueldmq | johnthetubaguy: I think we're talking the same thing (at least the same direction) all the time, but in different languages | 13:05 |
| samueldmq | johnthetubaguy: sdague I need to step in a meeting right now, talk to you later, thanks | 13:05 |
| johnthetubaguy | samueldmq: I think we are quite close here | 13:05 |
| alex_xu_ | danpb, dansmith, appreciate some help on https://review.openstack.org/#/c/168982/10, I updated it using configration file to define the abstact cpumodel, hope I can catch the last chance before freeze | 13:05 |
| *** sudipto has joined #openstack-nova | 13:06 | |
| *** moshele has joined #openstack-nova | 13:06 | |
| alex_xu_ | johnthetubaguy: ^ appreciate if you have time review that spec also :) | 13:06 |
| *** prashup has quit IRC | 13:07 | |
| johnthetubaguy | samueldmq: so the end user policy discoverability is the reason you need a sync I guess, but I keep thinking that needs to live in the projects, but thats a slightly separate conversation, potentially | 13:07 |
| *** coolsvap|away is now known as coolsvap | 13:08 | |
| johnthetubaguy | alex_xu_: can't promise right now, but thats surely an important one | 13:09 |
| *** changbl_ has quit IRC | 13:17 | |
| alex_xu_ | johnthetubaguy: I understand, thanks anyway! | 13:17 |
| *** elmiko has left #openstack-nova | 13:17 | |
| *** kylek3h has joined #openstack-nova | 13:17 | |
| *** jasondotstar has joined #openstack-nova | 13:17 | |
| *** scheuran has joined #openstack-nova | 13:17 | |
| *** openstack has quit IRC | 13:17 | |
| *** openstack has joined #openstack-nova | 13:18 | |
| -cameron.freenode.net- [freenode-info] if you're at a conference and other people are having trouble connecting, please mention it to staff: http://freenode.net/faq.shtml#gettinghelp | 13:18 | |
| *** _gryf is now known as _gryf|afk | 13:20 | |
| johnthetubaguy | bauzas: at least the freeze is getting folks to update their specs, but thats something for the nova-meeting on thursday I think, we need to keep folks productive, on both sides of the equation | 13:21 |
| bauzas | johnthetubaguy: yeah, my take is that it seems people are just afraid of missing the deadline, so are just trying to ping many people asking for reviews | 13:21 |
| ijuwang_away | back | 13:22 |
| *** ijuwang_away is now known as ijuwang | 13:22 | |
| johnthetubaguy | bauzas: true | 13:24 |
| *** abhishekk has quit IRC | 13:25 | |
| bauzas | to be clear, people asking for reviews related to the scheduler can drop their changes in https://etherpad.openstack.org/p/liberty-nova-priorities-tracking L47 | 13:27 |
| ndipanov | also worth noting that 5 or so ppl that are on the spec core combined with 2x +2s are a massive bottleneck | 13:27 |
| *** eharney has joined #openstack-nova | 13:27 | |
| *** mriedem has joined #openstack-nova | 13:27 | |
| ndipanov | all of this was decided by inertia and never re-examined | 13:27 |
| ndipanov | not to bring up all the other well known problems with the spec review process | 13:28 |
| johnthetubaguy | ndipanov: thats exactly why I want all of nova-core to have +2 on there, but not +W, but I was voted down, but we need to revisit that | 13:28 |
| ndipanov | I am not surprised at all that you were voted down | 13:29 |
| *** ajayaa has joined #openstack-nova | 13:29 | |
| mriedem | there was a vote? i just remember a lot of 'meh' and shoulder shrugging | 13:29 |
| *** aix has quit IRC | 13:29 | |
| danpb | johnthetubaguy: you could declare yourself el presidente non-benevolent dictator for life and change it :-) | 13:29 |
| *** prashup has joined #openstack-nova | 13:30 | |
| ndipanov | I feel that specs work well for some things | 13:31 |
| ndipanov | and very poorly for others | 13:31 |
| ndipanov | and the issues where they don't work is where we hit problems | 13:32 |
| *** kiwik-chenrui has quit IRC | 13:32 | |
| *** kiwik-chenrui has joined #openstack-nova | 13:32 | |
| *** sdake_ has joined #openstack-nova | 13:32 | |
| *** fawadkhaliq has joined #openstack-nova | 13:33 | |
| *** rlrossit has joined #openstack-nova | 13:33 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Add timing information for rootwrap daemon https://review.openstack.org/194602 | 13:34 |
| scheuran | Hi I'm trying to get a new vif_type approved for liberty. Is there any process for this, especially due to the os-vif-library planned? | 13:34 |
| scheuran | does this effort mean that new vif_types are totally blocked for liberty? | 13:34 |
| ndipanov | example - some specs by bauzas and around cells were extremely useful for me to understand the problem and think about it | 13:35 |
| johnthetubaguy | ndipanov: its normally us using specs poorly for some things | 13:35 |
| johnthetubaguy | ndipanov: where you can't design it up front you shouldn't, I have been pushing hard on that in my reviews | 13:35 |
| bauzas | agreed on the fact that we need to incrementally find a good cursor for saying if it needs a spec or not | 13:36 |
| *** artom has joined #openstack-nova | 13:36 | |
| dansmith | we've said "no need for a spec" more times this cycle than any others | 13:36 |
| bauzas | yup | 13:36 |
| johnthetubaguy | dansmith: +1 | 13:36 |
| dansmith | surely the incremental improvement plan applies to this process as well :) | 13:36 |
| mriedem | dansmith: johnthetubaguy: are specs needed for new vif types then? | 13:36 |
| mriedem | those are fairly mechanical, like new libvirt volume drivers | 13:37 |
| johnthetubaguy | mriedem: they are in my list of "stuck" specs right now | 13:37 |
| *** artom has quit IRC | 13:37 | |
| mriedem | scheuran: i'd get a bp and spec up today just in case the generic vif magic bullet doesn't happen | 13:37 |
| ndipanov | but I find it hard to argue these things as I am in general for a default-to-open approach :) | 13:37 |
| bauzas | tbc, I didn't want to open a can of worms, just try to see if we can lower the stress by saying "you miss the deadline ? no worries, we can still handle that if that deserves the project" | 13:37 |
| *** thangp has joined #openstack-nova | 13:37 | |
| ndipanov | well tbh that's what I understood the "open" exception process to be... | 13:38 |
| bauzas | please, don't see me trying to put the baby with the baby's bath water | 13:38 |
| bauzas | s/put/throw or whatever good English metaphor applies here | 13:39 |
| *** garyk has quit IRC | 13:39 | |
| *** fawadkhaliq has quit IRC | 13:39 | |
| scheuran | mriedem: thx, it's already up - just needs another round of review | 13:39 |
| openstackgerrit | wuhao proposed openstack/nova-specs: Add previous page links for Nova GET servers API https://review.openstack.org/165041 | 13:40 |
| *** jecarey has joined #openstack-nova | 13:41 | |
| *** doron_ has quit IRC | 13:41 | |
| markus_z | BobBall: About the "maintenance mode" thingy, I'll send an update on the mailing list. Would be great if you could have a look at it from a xen perspective. | 13:42 |
| *** aix has joined #openstack-nova | 13:42 | |
| *** tonytan4ever has joined #openstack-nova | 13:42 | |
| mriedem | scheuran: this is yours right? https://review.openstack.org/#/c/182280/ | 13:42 |
| scheuran | mriedm: yes | 13:43 |
| scheuran | mriedm: I guess the big question is if it can be approved in parallel to the planned os-vif-library stuff, or if I should update the text in order to depend on the new way that is proposed | 13:45 |
| *** vladikr has joined #openstack-nova | 13:45 | |
| mriedem | scheuran: do you have a link to os-vif-library? | 13:45 |
| mriedem | i see two other specs that were approved for liberty already for new vif types | 13:45 |
| mriedem | http://specs.openstack.org/openstack/nova-specs/specs/liberty/approved/vif-driver-ib-passthrough.html | 13:45 |
| mriedem | http://specs.openstack.org/openstack/nova-specs/specs/liberty/approved/vif-type-tap.html | 13:45 |
| scheuran | https://review.openstack.org/#/c/193668 | 13:46 |
| *** coolsvap is now known as coolsvap|away | 13:46 | |
| danpb | mriedem: oh hmm - i'd very much like us to refuse any vif types for liberty | 13:46 |
| danpb | until we have the os-vif stuff done | 13:47 |
| danpb | mriedem: url is https://review.openstack.org/#/c/193668/ | 13:47 |
| *** varunarya has joined #openstack-nova | 13:47 | |
| *** pczesno is now known as zzz_pczesno | 13:48 | |
| *** zzz_pczesno is now known as pczesno | 13:48 | |
| mriedem | so we're also holding the volume driver specs hostage until os-brick? https://review.openstack.org/#/c/181941/ and https://review.openstack.org/#/c/184295/ | 13:49 |
| scheuran | danpb: sure that would be the ideal thing, but in some cases it's just not possible to reuse an existing type due to different plug/unplug operations or to slightly different xml syntax | 13:49 |
| mriedem | scheuran: i think what danpb is saying is your new type goes in the os-vif-library | 13:50 |
| mriedem | then nova and neutron use the shared library | 13:50 |
| mriedem | scheuran: we have the same issue with volume drivers copied between cinder and nova | 13:50 |
| mriedem | and os-brick is supposed to solve that | 13:50 |
| ndipanov | johnthetubaguy, can we take a look at this - it's a bug fix, and it's stuck waiting for an opinion on some schema details that can easily be worked out in review | 13:50 |
| ndipanov | https://review.openstack.org/#/c/193576/ | 13:50 |
| *** kes has quit IRC | 13:51 | |
| *** ppal has joined #openstack-nova | 13:51 | |
| *** ppal has left #openstack-nova | 13:51 | |
| scheuran | mriedem: yes, once it's there. Would you propose to update the spec in order to depend on the new os-vif project? | 13:51 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add to_dict() to ImageMeta and ImageMetaProps objects https://review.openstack.org/186474 | 13:52 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add RequestSpec methods for primitiving into dicts https://review.openstack.org/187197 | 13:52 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Use RequestSpec object in HostManager https://review.openstack.org/191251 | 13:52 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Use the same pci_requests field for all filters and HostManager https://review.openstack.org/191250 | 13:52 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Remove unnecessary method in FilterScheduler https://review.openstack.org/188485 | 13:52 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Consider that all scheduler calls are IO Ops https://review.openstack.org/188486 | 13:52 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Create RequestSpec object https://review.openstack.org/145528 | 13:52 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Use RequestSpec object in the FilterScheduler https://review.openstack.org/188487 | 13:52 |
| bauzas | jaypipes: heya | 13:52 |
| danpb | scheuran: the whole point of the os-vif project is to decouple the plug/unplug operations from the vif type, precisely so we can reuse them | 13:52 |
| bauzas | jaypipes: so sounds like I have to ping some people for asking for reviews ;) | 13:52 |
| johnthetubaguy | ndipanov: its on my queue, but its a massive queue right now I am afraid | 13:52 |
| *** ssurana has joined #openstack-nova | 13:52 | |
| *** salv-orlando has joined #openstack-nova | 13:52 | |
| bauzas | jaypipes: because it's the general situation :) | 13:52 |
| *** agentle has joined #openstack-nova | 13:53 | |
| scheuran | danpb: right, I got that. The question is, how can I proceed with my new vif_type required? | 13:53 |
| bauzas | jaypipes: so given our lovely hangout, would appreciate if you could fire some comments for https://review.openstack.org/#/c/179224/ :) | 13:53 |
| mriedem | scheuran: it looks like your spec already calls out danpb's os-vif-library spec | 13:53 |
| mriedem | as a dependency | 13:53 |
| scheuran | danpb: should I update the spec and say, it will be implemented along the new approach (once approved)? | 13:54 |
| *** mpavone has joined #openstack-nova | 13:54 | |
| scheuran | mriedem: yes, but some details explain the part with the xml should still be implemented in the classical way | 13:54 |
| scheuran | but I could generalize this | 13:54 |
| *** moshele is now known as moshele_ | 13:55 | |
| mriedem | scheuran: have you reviewed https://review.openstack.org/#/c/193668/ and made sure it will work with what you need to do? | 13:55 |
| jaypipes | bauzas: yup, it's on my queue. | 13:55 |
| *** xyang1 has quit IRC | 13:56 | |
| *** electrocucaracha has joined #openstack-nova | 13:56 | |
| *** salv-orlando has quit IRC | 13:57 | |
| scheuran | mriedem: yes. Both directions would work. Getting the new vif_type in the old way, would not break this new feature. But I could also implement it in the new way | 13:57 |
| scheuran | mriedem: that would also work, if all the vif_types will be generalized like pointed out | 13:58 |
| scheuran | I already offered my help for bringing things forward.... | 13:58 |
| *** browne has joined #openstack-nova | 13:59 | |
| *** baoli has quit IRC | 13:59 | |
| *** rwsu has joined #openstack-nova | 14:01 | |
| *** otter768 has joined #openstack-nova | 14:01 | |
| *** agentle has quit IRC | 14:01 | |
| bauzas | jaypipes: coolio, so I can just drop from the channel, right ? :) | 14:01 |
| *** tpeoples has joined #openstack-nova | 14:02 | |
| *** bkopilov has joined #openstack-nova | 14:02 | |
| *** baoli has joined #openstack-nova | 14:02 | |
| ndipanov | johnthetubaguy, thanks anyway... | 14:03 |
| mriedem | scheuran: comments inline in your spec | 14:04 |
| mriedem | scheuran: i'm a bit confused about the security group thing | 14:04 |
| *** burt has joined #openstack-nova | 14:05 | |
| scheuran | mriedem: thx | 14:05 |
| *** otter768 has quit IRC | 14:05 | |
| sahid | mriedem, johnthetubaguy - any chance you remove -2 from this serie https://review.openstack.org/#/c/185910/ | 14:07 |
| mriedem | sahid: the bp isn't approved | 14:07 |
| *** xgsa has quit IRC | 14:08 | |
| mriedem | oh i see the spec is though https://review.openstack.org/#/c/187509/ | 14:08 |
| mriedem | sahid: removed my -2 | 14:08 |
| sahid | cool :) | 14:08 |
| *** zz_jgrimm is now known as jgrimm | 14:09 | |
| *** jmccrory has quit IRC | 14:10 | |
| *** busterswt has joined #openstack-nova | 14:11 | |
| *** jmccrory has joined #openstack-nova | 14:11 | |
| *** sdake_ has quit IRC | 14:12 | |
| *** xgsa has joined #openstack-nova | 14:13 | |
| *** varunarya has quit IRC | 14:14 | |
| *** artom has joined #openstack-nova | 14:15 | |
| *** marun has joined #openstack-nova | 14:16 | |
| *** varunarya has joined #openstack-nova | 14:16 | |
| *** mgarza_ has quit IRC | 14:17 | |
| *** karimb has quit IRC | 14:17 | |
| *** iamjarvo has joined #openstack-nova | 14:19 | |
| larsks | 587373 | 14:19 |
| dansmith | larsks: I disagree | 14:20 |
| dansmith | larsks: I think it's 408185 | 14:20 |
| *** vladikr has quit IRC | 14:21 | |
| *** marun has quit IRC | 14:22 | |
| *** karimb has joined #openstack-nova | 14:24 | |
| markus_z | johnthetubaguy: Would you please remove your -2 from https://review.openstack.org/#/c/188058/ The spec is pushed. | 14:25 |
| markus_z | ? | 14:25 |
| *** dboik has joined #openstack-nova | 14:25 | |
| *** VW_ has quit IRC | 14:26 | |
| mriedem | markus_z: the spec isn't approved | 14:26 |
| mriedem | markus_z: the -2 is on your nova change until the spec is approved | 14:26 |
| mriedem | https://review.openstack.org/#/c/194222/ | 14:26 |
| markus_z | mriedem: ah, ok, wasn't sure about that | 14:26 |
| markus_z | It hit me by surprise that a spec was needed for that. Sounded trivial to me. | 14:27 |
| *** electroc_ has joined #openstack-nova | 14:27 | |
| *** electrocucaracha has quit IRC | 14:28 | |
| *** nkrinner has quit IRC | 14:29 | |
| mriedem | hemnafk: with the os-brick change, will we still need libvirt volume drivers to do the get_config call? i see that most of the LibvirtFibreChannelVolumeDriver is removed in https://review.openstack.org/#/c/175569/ but get_config is still there | 14:29 |
| *** jasondotstar has quit IRC | 14:31 | |
| *** gszasz has quit IRC | 14:31 | |
| *** nihilifer has joined #openstack-nova | 14:31 | |
| mriedem | ha https://review.openstack.org/#/c/194412/ | 14:33 |
| bauzas | johnthetubaguy: oh btw. I think https://review.openstack.org/#/c/181054/ isn't necessary since it's only needing 2 conf flags and the implem seems trivial. Your opinion ? | 14:33 |
| mriedem | bug as feature https://review.openstack.org/#/c/194412/ | 14:33 |
| bauzas | (I know that you're busy, I don't want a review, just an opinion saying if modifying a filter needs a spec) | 14:33 |
| bauzas | MHO is that it's not needed | 14:34 |
| *** ctrath has quit IRC | 14:34 | |
| openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Support --all-tenants in floatingips-list https://review.openstack.org/171389 | 14:34 |
| *** adrian_otto has joined #openstack-nova | 14:35 | |
| artom | johnthetubaguy, melwitt, and anyone else who's into reviewing API specs, I think I've addressed the concerns about pagination in the above, can I ask you for some more reviews on the above | 14:35 |
| lxsli | gilliard: ^^ | 14:35 |
| *** vladikr has joined #openstack-nova | 14:36 | |
| artom | I'm not normally this pushy, but the deadline is approaching ;) | 14:36 |
| *** afazekas has quit IRC | 14:36 | |
| jaypipes | bauzas: lol | 14:38 |
| kashyap | markus_z: Your nice descriptive email is buried under quoted text. :-) | 14:38 |
| markus_z | kashyap: Oh, did I screw it? | 14:39 |
| bauzas | kashyap: my layout is fine | 14:39 |
| kashyap | markus_z: Nah, layout is good. | 14:39 |
| markus_z | I thought appending the new stuff at the bottom of a mail is what most people do. | 14:40 |
| kashyap | It appeared as a new thread to me. | 14:40 |
| *** adrian_otto has quit IRC | 14:41 | |
| kashyap | Please ignore my bike-shedding on that. Just that your propose a solution, you could have added a "proposal" to the subject. | 14:41 |
| lxsli | I haven't seen the email in question but... when bottom-quoting, I wish people would snip more aggressively | 14:41 |
| markus_z | kashyap: Ah, that's my fault, I didn't have the original mail and I had to copy&paste from the internet | 14:41 |
| lxsli | sometimes I have to scroll past three pages of quotes to see "+1", it's annoying | 14:42 |
| *** gszasz has joined #openstack-nova | 14:42 | |
| jaypipes | johnthetubaguy, danpb: https://review.openstack.org/#/c/128825/ from vladikr looks good to go now. | 14:42 |
| kashyap | markus_z: Just that these days the 140-character attention is the rage, and I slightly wondered if people concerend will read through. | 14:43 |
| jaypipes | kashyap: damn that Twitter. | 14:43 |
| kashyap | lxsli: If you're using `mutt` for mail reader, there's an excellent little trick to rescue: toggle-quoted command ('T' by default) | 14:44 |
| *** moshele_ has quit IRC | 14:44 | |
| kashyap | jaypipes: :-) | 14:44 |
| *** xyang1 has joined #openstack-nova | 14:44 | |
| *** adrian_otto has joined #openstack-nova | 14:44 | |
| *** gszasz has quit IRC | 14:44 | |
| *** gszasz has joined #openstack-nova | 14:45 | |
| bauzas | nova-scheduler meeting in 15 mins on #openstack-meeting | 14:45 |
| *** gszasz has quit IRC | 14:45 | |
| bauzas | folks having questions or review requests re: to the scheduler can join | 14:45 |
| *** jasondotstar has joined #openstack-nova | 14:46 | |
| markus_z | kashyap: hm, yeah, maybe it I'll put it on top the next time. I just make the changes I suggested and wait for the yelling :) | 14:46 |
| *** thedodd has joined #openstack-nova | 14:47 | |
| *** gszasz has joined #openstack-nova | 14:47 | |
| *** adrian_otto1 has joined #openstack-nova | 14:47 | |
| *** mgarza_ has joined #openstack-nova | 14:48 | |
| *** heyongli has quit IRC | 14:49 | |
| *** adrian_otto has quit IRC | 14:49 | |
| *** adrian_otto has joined #openstack-nova | 14:50 | |
| edleafe | it's always best to trim the quoted text to just the section you're replying to. | 14:50 |
| *** heyongli has joined #openstack-nova | 14:50 | |
| jroll | jaypipes: hey, I pushed up that spec from yesterday's conversation https://review.openstack.org/#/c/194453/2 | 14:51 |
| *** sdake has joined #openstack-nova | 14:51 | |
| jaypipes | jroll: rock on, thank you. I'll review it shortly. | 14:51 |
| *** kiran-r has joined #openstack-nova | 14:51 | |
| *** adrian_otto1 has quit IRC | 14:52 | |
| *** adrian_otto1 has joined #openstack-nova | 14:52 | |
| jroll | jaypipes: cool, thanks in advance :) | 14:52 |
| *** erlon has joined #openstack-nova | 14:52 | |
| *** agireud has joined #openstack-nova | 14:53 | |
| jaypipes | jroll: -2 | 14:53 |
| jaypipes | jroll: just kiddin ;) | 14:53 |
| jroll | :) | 14:53 |
| dansmith | johnthetubaguy: I propose we drop this spec, unblock ildikov, and do the spec for the big work in M as proposed: https://review.openstack.org/#/c/193510/ | 14:53 |
| *** IanGovett has quit IRC | 14:53 | |
| *** tojuvone has quit IRC | 14:54 | |
| lxsli | kashyap: hehe yes I know it thanks - but I don't know ahead of time when someone will do it | 14:54 |
| *** adrian_otto has quit IRC | 14:54 | |
| *** adrian_otto1 has quit IRC | 14:55 | |
| lxsli | kashyap: I do sometimes resort to that for the ridiculous threads but I usually just stop reading them | 14:55 |
| lxsli | kashyap: also, mutt for the win \o/ | 14:55 |
| *** VW_ has joined #openstack-nova | 14:55 | |
| *** mleroy_ has joined #openstack-nova | 14:56 | |
| ildikov | dansmith: thanks for the proposal | 14:57 |
| dansmith | ildikov: sorry you had to write a spec to show it, but I'm thankful you did.. helps make it clear | 14:58 |
| ildikov | dansmith: johnthetubaguy: I'm fine with doing the first phase without a spec if there is an agreement on this and it is documented | 14:58 |
| *** zzzeek has joined #openstack-nova | 14:58 | |
| *** adrian_otto has joined #openstack-nova | 14:58 | |
| *** yamahata has joined #openstack-nova | 14:58 | |
| *** IanGovett has joined #openstack-nova | 14:59 | |
| johnthetubaguy | dansmith: thats cool, seems like the bits left are a spec-less blueprint, with the other bit for M | 14:59 |
| dansmith | johnthetubaguy: could do that too | 14:59 |
| ildikov | dansmith: np, it is important to be on the same page, so it's worth it | 14:59 |
| *** unicell has joined #openstack-nova | 15:01 | |
| dansmith | johnthetubaguy: ah, did you just approve that bp? | 15:01 |
| johnthetubaguy | ildikov: dansmith: so my take is, we don't need a blueprint for your patch, but its handy to track things, so I have just approved the BP: https://blueprints.launchpad.net/nova/+spec/optional-on-shared-storage-flag-in-rebuild-instance | 15:01 |
| dansmith | ah, looks like so | 15:01 |
| johnthetubaguy | dansmith: yeah, I figured we don't need one, but it could be handy due to the cross release thingy | 15:02 |
| *** adrian_otto1 has joined #openstack-nova | 15:02 | |
| *** kiran-r has quit IRC | 15:02 | |
| dansmith | cool, so ildikov please drop the first spec and go for the code :) | 15:02 |
| *** ijw_ has joined #openstack-nova | 15:02 | |
| dansmith | johnthetubaguy: and it was already created | 15:02 |
| *** matrohon has joined #openstack-nova | 15:02 | |
| johnthetubaguy | dansmith: exactly | 15:02 |
| *** unicell1 has quit IRC | 15:02 | |
| *** adrian_otto1 has quit IRC | 15:02 | |
| *** ctrath has joined #openstack-nova | 15:03 | |
| *** adrian_otto has quit IRC | 15:03 | |
| *** artom has quit IRC | 15:03 | |
| bauzas | last call for boarding on nova-scheduler meeting at #openstack-meeting | 15:03 |
| ildikov | johnthetubaguy: dansmith: ok, cool, tnx | 15:04 |
| openstackgerrit | Andreas Scheuring proposed openstack/nova-specs: Spec to Add 'macvtap' as vif type to novas libvirt driver. https://review.openstack.org/182280 | 15:04 |
| ildikov | johnthetubaguy: dansmith: we will abandon the review and refer to the Launchpad BP | 15:04 |
| dansmith | ildikov: cool, thanks | 15:04 |
| johnthetubaguy | ildikov: perfect | 15:04 |
| edleafe | lxsli: jaypipes: scheduler meeting now | 15:04 |
| *** bauzas is now known as JustinBieber | 15:04 | |
| johnthetubaguy | ildikov: we can keep that backlog spec going, they are not affected by the freeze, FWIW, at least thats my plan A | 15:05 |
| mriedem | gmann: what's up with this comment about all_tenants in the v2.1 servers API? http://git.openstack.org/cgit/openstack/nova/tree/nova/api/openstack/compute/plugins/v3/servers.py#n331 | 15:05 |
| mriedem | gmann: it says to remove the commented out code at some point - can we do that? | 15:05 |
| *** JustinBieber is now known as NotJustinBieber | 15:05 | |
| scheuran | mriedem: I just updated the spec regarding macvtap-vif | 15:05 |
| mriedem | gmann: i'm wondering if server list --all-tenants is basically not enforced today | 15:05 |
| mriedem | for v2.1 | 15:05 |
| ildikov | dansmith: johnthetubaguy: can we do smth with the service status change notification one: https://review.openstack.org/#/c/182350/ ? It will be delayed by two releases if we don't have it done in L | 15:06 |
| *** NotJustinBieber is now known as bauzas | 15:06 | |
| johnthetubaguy | ildikov: why will it be delated by two releases? | 15:06 |
| ildikov | johnthetubaguy: IIUC the second part of the spec has to land in M in order to follow the upgrade path | 15:06 |
| *** vilobhmm has joined #openstack-nova | 15:06 | |
| *** claudiub has quit IRC | 15:07 | |
| *** garyk has joined #openstack-nova | 15:07 | |
| ildikov | johnthetubaguy: I meant that most probably it can land in N the earliest as hopefully the notification tidy up work will take place in M | 15:07 |
| dansmith | ildikov: I'll have to look at that one | 15:07 |
| *** r-daneel has joined #openstack-nova | 15:07 | |
| scheuran | mriedem: it's now written in a more generic way, mentioning the os-vif-library as way to implement it if it will come | 15:07 |
| dansmith | ildikov: let us continue on our queues | 15:07 |
| *** atuvenie__ has joined #openstack-nova | 15:08 | |
| johnthetubaguy | ildikov: now I see what you are saying, if we can get help fixing notifications now, that would be much quicker, but yeah, let me think about all that | 15:08 |
| ildikov | dansmith: sure, sorry :) I will be available for a while in case of anything pops up, like for the onSharedStorage one | 15:09 |
| mriedem | alex_xu_: can you reivew this when you get a chance? https://review.openstack.org/#/c/171389/ | 15:09 |
| *** claudiub has joined #openstack-nova | 15:10 | |
| ildikov | johnthetubaguy: I quickly discussed it with dansmith, so we will help as much as we can in that process, but it seems that in L we can only get to figure out a design, but most likely the implementation will fall into M the earliest | 15:10 |
| edleafe | mriedem: alex_xu_ has a new daughter today. He might not be around much. :) | 15:10 |
| mriedem | sorry i need to see a doctor's note | 15:11 |
| dansmith | ildikov: well, that was my estimation of achievable bits based on inertia | 15:11 |
| *** atuvenie_ has quit IRC | 15:11 | |
| *** sudipto has quit IRC | 15:11 | |
| *** varunarya has quit IRC | 15:12 | |
| *** vilobhmm has quit IRC | 15:12 | |
| dansmith | vladikr: https://review.openstack.org/#/c/128825/12 | 15:12 |
| dansmith | vladikr: am I nuts? | 15:12 |
| ildikov | dansmith: considering that we are in the Liberty release and the notification tidy up process has been discussed several times for a very long while now in different levels that estimation sounds reasonable | 15:13 |
| * dansmith waits for someone to compliment his discerning eye | 15:13 | |
| vladikr | dansmith, reading.. | 15:13 |
| *** READ10 has joined #openstack-nova | 15:14 | |
| vladikr | dansmith, I think I don't see what you see.. | 15:14 |
| dansmith | okay, let me pull it down | 15:15 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: libvirt: Always default device names at boot https://review.openstack.org/189632 | 15:15 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: virt: add get_device_name_for_instance to the base driver class https://review.openstack.org/180637 | 15:15 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: libvirt: implement get_device_name_for_instance https://review.openstack.org/180638 | 15:15 |
| johnthetubaguy | ildikov: so I am looking at this a bit differently, we need some new folks to step up and help drive the notification, because the current folks are currently busy till at least the end of L with the current more/equally important work, so if we can get some people to help with the coding, it could happen quicker | 15:16 |
| dansmith | vladikr: jaypipes: danpb: http://pastebin.com/1WAthCDg | 15:17 |
| dansmith | vladikr: unicode character 0xd0b0 | 15:17 |
| danpb | dansmith: ? | 15:17 |
| dansmith | danpb: see my comment on https://review.openstack.org/#/c/128825 | 15:17 |
| danpb | everything looks normal when i view it | 15:18 |
| dansmith | http://www.fileformat.info/info/unicode/char/0430/index.htm | 15:18 |
| *** dsanders has joined #openstack-nova | 15:18 | |
| dansmith | depends on how your platform renders that ^ | 15:18 |
| danpb | just like a normal letter a | 15:19 |
| ildikov | johnthetubaguy: I haven't seen a spec for it yet. is there any? | 15:19 |
| dansmith | danpb: it's not ascii though, it's a two-byte unicode a | 15:19 |
| *** gszasz has quit IRC | 15:20 | |
| danpb | i guess my fonts just happen to make it look the same as an ascii a | 15:20 |
| johnthetubaguy | ildikov: thats the next required step, get someone to write up a spec (this is one of the post summit follow up steps that I requested help for on the ML, I don't think anyone has stepped up yet) | 15:20 |
| vladikr | dansmith, okay.. let me resend it.. | 15:20 |
| *** belmoreira has quit IRC | 15:21 | |
| ildikov | johnthetubaguy: I cannot promise anything until have an estimation about the amount of work needed | 15:21 |
| johnthetubaguy | ildikov: its basically use olso.versionedobjects to model the notification payloads, and add versioning, its a relatively large chunk of work | 15:21 |
| dansmith | johnthetubaguy: and support both formats for a while during a transition... | 15:22 |
| dansmith | (somehow) | 15:22 |
| ildikov | johnthetubaguy: yeap, I heard about versionedobjects, would need to take a deeper look though | 15:23 |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova-specs: libvirt: virtio-net multiqueue https://review.openstack.org/128825 | 15:23 |
| *** kiran-r has joined #openstack-nova | 15:23 | |
| ildikov | johnthetubaguy: dansmith: I think writing up a spec would be good to start by one of the current folks | 15:23 |
| dansmith | vladikr: +W, thanks | 15:24 |
| johnthetubaguy | ildikov: agreed, just they are all super busy right now | 15:24 |
| vladikr | dansmith, thanks | 15:24 |
| ildikov | johnthetubaguy: dansmith: in order to consider all the aspects, etc. | 15:24 |
| dansmith | vladikr: if you look at gerrit's diff you'll see I'm not the only one that can tell the difference :P | 15:24 |
| johnthetubaguy | ildikov: its a todo list I am tracking, but its not going to be quick if folks can't step up an help, thats the crux of it | 15:24 |
| vladikr | dansmith, yeah, I've seen it now | 15:24 |
| ildikov | johnthetubaguy: I got that part, this is why I menioned the estimation we had with dansmith as it sounds quite reasonable, even if it's not exactly what we all would desire | 15:25 |
| *** mpaolino has quit IRC | 15:25 | |
| *** amotoki has quit IRC | 15:26 | |
| ildikov | johnthetubaguy: it seems like a chicken-egg problem from the point of view that until we don't have a spec to start with we will not have folks to join, but until we don't have folks who could join we don't have the bandwidth to start | 15:26 |
| *** rajesht has quit IRC | 15:26 | |
| dansmith | ildikov: right, infinite bandwidth is not a reasonable expectation | 15:26 |
| dansmith | some things have to come after others and in reality, the versioning of notifications and your service notifications are fairly low on the priority list | 15:27 |
| *** rotbeard has joined #openstack-nova | 15:27 | |
| *** changbl_ has joined #openstack-nova | 15:28 | |
| ildikov | johnthetubaguy: to avoid misunderstandings, I'm totally on your side, but I cannot volunteer to write up a spec, as even if I would have the bandwidth it would take much time and I'm sure that wouldn't be that useful on the first place maybe not even on the second... :S | 15:28 |
| ildikov | dansmith: service notifications are not a big change and does not really make it harder to do the versioning | 15:29 |
| ildikov | dansmith: as much as I can see | 15:29 |
| dansmith | ildikov: the more notifications we have, the more compatibility we have to maintain | 15:29 |
| dansmith | ildikov: agree it's not massive, nor did I ever say that one has to follow the other necessarily | 15:30 |
| *** jwcroppe_ has quit IRC | 15:30 | |
| dansmith | ildikov: but on the scale of relative importance, service notifications are not critical, IMHO | 15:30 |
| *** jwcroppe has joined #openstack-nova | 15:30 | |
| *** kfox1111 has joined #openstack-nova | 15:31 | |
| *** matrohon has quit IRC | 15:32 | |
| johnthetubaguy | ildikov: understood | 15:32 |
| *** gszasz has joined #openstack-nova | 15:32 | |
| ildikov | dansmith: I agree that it's not the most critical item on the list | 15:32 |
| *** armax has joined #openstack-nova | 15:33 | |
| *** mestery has joined #openstack-nova | 15:34 | |
| dansmith | ildikov: just marked the M part of the evacuate spec as -2 to avoid confusion about it landing in M | 15:34 |
| *** amotoki has joined #openstack-nova | 15:35 | |
| *** Flcn has joined #openstack-nova | 15:35 | |
| dansmith | ildikov: when M specs open, move it into that directory and ping me to drop the -2 | 15:35 |
| *** jwcroppe has quit IRC | 15:35 | |
| ildikov | dansmith: but seem more imprtant than delaying it for an undefined amount of time | 15:35 |
| *** mestery_ has joined #openstack-nova | 15:35 | |
| dansmith | ildikov: well, that's the infinite bandwidth problem, but you can certainly make the argument that it's worth doing before the refactor | 15:35 |
| ildikov | dansmith: sure, thanks for the administartion on it | 15:35 |
| dansmith | like I said, I'm not entirely opposed... | 15:35 |
| Flcn | Hello! Does anybody here use more than 2 external neutron network per VM? | 15:36 |
| *** kiranr has joined #openstack-nova | 15:36 | |
| ildikov | dansmith: well, it is a small scope change, I don't know how much we can do it in a way to prepare for the refactor too, but if there is a chance for that, then we are open to follow that way | 15:38 |
| *** matrohon has joined #openstack-nova | 15:38 | |
| mriedem | melwitt: can you link me to your low hanging fruit etherpad again? | 15:38 |
| *** mestery has quit IRC | 15:39 | |
| ildikov | johnthetubaguy: do you have the link to the help request mail? | 15:39 |
| *** mestery_ is now known as mestery | 15:40 | |
| *** kiran-r has quit IRC | 15:40 | |
| *** destiny has quit IRC | 15:40 | |
| *** haypo has quit IRC | 15:41 | |
| *** sdake has quit IRC | 15:42 | |
| kfox1111 | Hi all. I'm looking to get some more nova folks taking a look at: https://review.openstack.org/#/c/186617/ Please take a look if you have a minute. | 15:42 |
| *** mgarza_ has quit IRC | 15:42 | |
| *** marun has joined #openstack-nova | 15:43 | |
| *** mgarza_ has joined #openstack-nova | 15:43 | |
| ildikov | mriedem: https://etherpad.openstack.org/p/nova-low-hanging-fruit | 15:45 |
| ndipanov | lxsli, interesting https://review.openstack.org/#/c/84906/19/specs/juno/persistent-resource-claim.rst | 15:45 |
| ndipanov | but this is merged for juno | 15:46 |
| ndipanov | what about L | 15:46 |
| *** lyarwood has quit IRC | 15:46 | |
| lxsli | ndipanov: yes I'd like to revive that in Muppet or Narnia | 15:46 |
| *** jwcroppe has joined #openstack-nova | 15:46 | |
| *** lyarwood has joined #openstack-nova | 15:46 | |
| ndipanov | hmm | 15:46 |
| *** pbrooko has quit IRC | 15:47 | |
| ndipanov | let me think about it | 15:47 |
| lxsli | ndipanov: it has to queue behind resource-objects, request-spec, bauzas said your HostState stuff as well | 15:47 |
| *** mgarza_ has quit IRC | 15:47 | |
| ndipanov | I might do some things differently in that case | 15:47 |
| *** atuvenie__ has quit IRC | 15:47 | |
| johnthetubaguy | ildikov: I not sure where that is right now, the key bit is this etherpad: https://etherpad.openstack.org/p/YVR-nova-liberty-summit-action-items | 15:47 |
| lxsli | ndipanov: yeah it will need to be re-proposed and updated I'm sure | 15:47 |
| mriedem | ildikov: thanks | 15:48 |
| ndipanov | well so here's the thing about this | 15:48 |
| mriedem | rlrossit: fyi https://etherpad.openstack.org/p/nova-low-hanging-fruit | 15:48 |
| ndipanov | currently instead of having a claim object | 15:48 |
| bauzas | ndipanov: that's what we agreed on like 10 mins ago :) | 15:48 |
| *** dsanders has quit IRC | 15:48 | |
| ildikov | johnthetubaguy: ok, thanks | 15:49 |
| ndipanov | we just consume those resources using the instance | 15:49 |
| *** yassine_ has quit IRC | 15:49 | |
| ndipanov | so this also goes back to the story we never really figured out | 15:49 |
| ildikov | mriedem: np | 15:49 |
| ndipanov | why do we need request_spec, instance, claim, filter_specs... all of which keep data about the same thing more or less | 15:49 |
| ndipanov | it just becomes very difficult to program against | 15:50 |
| *** armax has quit IRC | 15:50 | |
| ndipanov | so I'd like to sort that out first | 15:51 |
| ndipanov | but I'm realistic | 15:51 |
| bauzas | for the moment, resource-objects doesn't pretend to be persisted :) | 15:51 |
| *** marun has quit IRC | 15:51 | |
| ndipanov | bauzas, yeah I am not sure how I feel about it | 15:51 |
| *** mgarza_ has joined #openstack-nova | 15:51 | |
| ndipanov | I think that the fact that resource object and filter specs are leaked into compute hosts | 15:52 |
| johnthetubaguy | bauzas: is someone still working on the overall vision thing for this stuff? the devref doc? | 15:52 |
| ndipanov | is a critical design flaw | 15:52 |
| bauzas | ndipanov: what do you mean by filter specs ? | 15:52 |
| lxsli | ndipanov: requests and claims are quite different to me - one is abstract (some CPUs), the other is concrete (CPUs 2+3) | 15:52 |
| lxsli | ndipanov: but I agree some of the other stuff seems redundant | 15:52 |
| ndipanov | lxsli, true | 15:52 |
| ndipanov | I'm not saying that we should not have them, just that we should have a clear picture and a programming model | 15:53 |
| johnthetubaguy | we need to store the user request separate to what they got, thats not quite the same thing | 15:53 |
| johnthetubaguy | like the scheduler hints, etc | 15:53 |
| bauzas | ndipanov: that's the main difference between the RequestSpec and any resource object | 15:53 |
| johnthetubaguy | after a resize, its interesting to know its not what they requested at first, I guess | 15:53 |
| ndipanov | johnthetubaguy, currently it's all kind of on the instance but not really | 15:53 |
| ndipanov | with a bunch of stuff that is not | 15:53 |
| *** fawadkhaliq has joined #openstack-nova | 15:53 | |
| bauzas | ndipanov: the RequestSpec is there for keeping a trace of what the user asked | 15:53 |
| bauzas | ndipanov: not what the instance is about, or what it became | 15:54 |
| *** electroc_ is now known as electrocucaracha | 15:54 | |
| bauzas | johnthetubaguy: devref for what ? the big scheduler changes ? that's on my due :( | 15:54 |
| ndipanov | but currently it leaks into the compute service | 15:54 |
| bauzas | ndipanov: agreed, that's another problem we need to solve | 15:55 |
| bauzas | ndipanov: and see why it's needed on the compute side | 15:55 |
| *** marun has joined #openstack-nova | 15:55 | |
| ndipanov | I don't think it is, it's just used for the retry filter spec | 15:56 |
| jaypipes | dansmith: what am I looking at in that paste above? | 15:56 |
| *** beagles is now known as beagles_brb | 15:56 | |
| bauzas | ndipanov: okay, lemme take some time for code reviewing, it deserves your question | 15:56 |
| dansmith | jaypipes: proof that vladikr had a unicode 'a' in there instead of the ascii one | 15:57 |
| jaypipes | dansmith: oooohhh... | 15:57 |
| dansmith | jaypipes: it's no longer relevant though, as it's been fixed :) | 15:57 |
| ndipanov | fwiw - I think the instance should be the central object | 15:57 |
| jaypipes | ndipanov: fwiw, I think the nova.objects.DanSmith should be the central object. | 15:57 |
| *** rwsu has quit IRC | 15:58 | |
| ndipanov | jaypipes, that's an awesome april 1st commit :D | 15:58 |
| dansmith | jaypipes: oh man, I'd hate to see the methods people add to that object... | 15:58 |
| *** ZZelle is now known as ZZelle-away | 15:58 | |
| bauzas | ndipanov: um, you're right, we're just passing request_spec over the wire but not reading it - fantastic | 15:58 |
| *** unicell has quit IRC | 15:59 | |
| jaypipes | dansmith: nova.objects.DanSmith.get_off_my_lawn(). | 15:59 |
| ndipanov | bauzas, but we read filter_spec | 15:59 |
| bauzas | ndipanov: and it seems filt_props is only given for the retry dict | 15:59 |
| *** dboik_ has joined #openstack-nova | 15:59 | |
| bauzas | ndipanov: yup | 15:59 |
| ndipanov | or properties or whatever it's called | 15:59 |
| *** dboik_ has quit IRC | 15:59 | |
| *** sahid has quit IRC | 15:59 | |
| bauzas | ndipanov: filter_properties indeed | 15:59 |
| ndipanov | jaypipes, nova.objects.DanSmith.get_off_my_lawn(*kids) | 15:59 |
| *** dboik_ has joined #openstack-nova | 15:59 | |
| dansmith | ndipanov: +2 | 15:59 |
| jaypipes | ndipanov: ! yes :) | 16:00 |
| jaypipes | nova.objects.DanSmith.avoid_humans(*mock.ANY) | 16:00 |
| *** chen12 has quit IRC | 16:00 | |
| edleafe | nova.objects.DanSmith.no(your_proposal) | 16:00 |
| jaypipes | :) | 16:00 |
| jaypipes | lmao | 16:00 |
| *** hemnafk is now known as hemna | 16:01 | |
| ndipanov | hahaha | 16:01 |
| dansmith | heh | 16:01 |
| *** chen12 has joined #openstack-nova | 16:01 | |
| jaypipes | dansmith: we are quickly rounding out your object interface here. | 16:01 |
| dansmith | lol | 16:01 |
| bauzas | ndipanov: so given that, I think that's fine - provided any call to select_dest() is reading the Spec object and persisting it back, we no longer need to pass either filt_props or request_sepc | 16:01 |
| *** otter768 has joined #openstack-nova | 16:02 | |
| scheuran | Can anyone please take the macvtap-vif spec on its review agenda? It's updated to be either implemented along the proposed os-vif-library way or in the classical way, if os-vif will not make it. (https://review.openstack.org/#/c/182280). Thx | 16:02 |
| bauzas | ndipanov: we're just passing them because we don't persist the retry dict | 16:02 |
| ndipanov | def is_approved(proposal): return False | 16:02 |
| ndipanov | bauzas, well not sure retries need to be persisted | 16:03 |
| *** dboik has quit IRC | 16:03 | |
| ndipanov | but yeah that whole intreaction needs a rework | 16:03 |
| ndipanov | that;s basically what I meant by fixing the scheduler interface | 16:03 |
| *** jgrimm is now known as zz_jgrimm | 16:03 | |
| danpb | dansmith: would like your opinion on this, since it is talking about objectifying the VIF type config data and objectification is your speciality https://review.openstack.org/#/c/193668/ | 16:03 |
| dansmith | danpb: yeah, I was trying to avoid it, but I should probably dig in :/ | 16:04 |
| *** armax has joined #openstack-nova | 16:05 | |
| *** sdake has joined #openstack-nova | 16:05 | |
| markus_z | anyone from hyper-v around? | 16:05 |
| claudiub | markus_z: hi. | 16:06 |
| *** arnaud____ has joined #openstack-nova | 16:06 | |
| *** READ10 has quit IRC | 16:06 | |
| markus_z | claudiub: Hi, I'm not sure if you followed the discussion about the "evacuate" vs. "host maintenance mode". | 16:06 |
| markus_z | claudiub: http://lists.openstack.org/pipermail/openstack-dev/2015-June/067787.html | 16:07 |
| *** otter768 has quit IRC | 16:07 | |
| markus_z | claudiub: I think hyper-v was mislead here too | 16:07 |
| johnthetubaguy | markus_z: I should totally write up all the different things, I have a feeling I started something on that somewhere | 16:07 |
| dansmith | mriedem: are you looking at danpb's of-vif thing? you probably should, given it's relation to brick | 16:08 |
| *** alex_klimov has quit IRC | 16:08 | |
| *** armax has quit IRC | 16:08 | |
| dansmith | er, os-vif | 16:08 |
| mriedem | dansmith: it's living comfortably in a browser tab | 16:08 |
| mriedem | currently reaping the results of https://review.openstack.org/#/c/192348/ | 16:09 |
| mriedem | sdague: i have the race figured out for bug 1249065 now | 16:10 |
| openstack | bug 1249065 in OpenStack Compute (nova) "Nova throws 400 when attempting to add floating ip (instance.info_cache.network_info is empty)" [Medium,Confirmed] https://launchpad.net/bugs/1249065 | 16:10 |
| mriedem | given that extra logging | 16:10 |
| mriedem | \o/ | 16:10 |
| markus_z | claudiub: The bottom line is, I think hyper-v's checkmark here is a mistake: http://docs.openstack.org/developer/nova/support-matrix.html#operation_evacuate | 16:12 |
| *** achanda has joined #openstack-nova | 16:12 | |
| markus_z | I'm preparing a patch for s390 and I would add hyper-v too if you can confirm this. | 16:13 |
| *** nihilifer has quit IRC | 16:13 | |
| *** romainh has left #openstack-nova | 16:13 | |
| *** veena has joined #openstack-nova | 16:13 | |
| claudiub | markus_z: yeah. got it. I will check it out. Add me to cc to it please. | 16:13 |
| *** iamjarvo has quit IRC | 16:13 | |
| claudiub | markus_z: thanks for bringing it up. :) | 16:14 |
| *** tonytan4ever has quit IRC | 16:17 | |
| *** ijw_ has quit IRC | 16:17 | |
| *** scheuran has quit IRC | 16:17 | |
| *** nic has joined #openstack-nova | 16:19 | |
| *** mpavone has quit IRC | 16:20 | |
| *** samueldmq has quit IRC | 16:21 | |
| *** annashen has joined #openstack-nova | 16:22 | |
| *** marun has quit IRC | 16:23 | |
| *** salv-orlando has joined #openstack-nova | 16:23 | |
| *** samueldmq has joined #openstack-nova | 16:24 | |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: hypervisor support matrix: fix "evacuate" for s390 and hyper-v https://review.openstack.org/194730 | 16:24 |
| *** yamahata has quit IRC | 16:24 | |
| openstackgerrit | Earle F. Philhower, III proposed openstack/nova: Add os_brick-based VolumeDriver for HGST connector https://review.openstack.org/194412 | 16:26 |
| jaypipes | mriedem, dansmith: you'll be delighted to hear that Jules has just ordered diapers for one of the pugs. | 16:27 |
| dansmith | jaypipes: lol | 16:27 |
| dansmith | jaypipes: pictures please, when available | 16:27 |
| jaypipes | dansmith: but of course. | 16:28 |
| *** _kiran_ has joined #openstack-nova | 16:28 | |
| *** mestery has quit IRC | 16:28 | |
| jaypipes | johnthetubaguy: are you around? did you catch my request from yesterday about snikitin's server instance tagging spec? | 16:29 |
| *** lucasagomes is now known as lucas-dinner | 16:29 | |
| *** krtaylor has quit IRC | 16:29 | |
| *** markus_z has quit IRC | 16:29 | |
| *** dims has quit IRC | 16:30 | |
| hemna | mriedem, ping | 16:31 |
| *** kiranr has quit IRC | 16:32 | |
| johnthetubaguy | jaypipes: I am around, I haven't got back around to review that one yet | 16:32 |
| *** dims has joined #openstack-nova | 16:33 | |
| johnthetubaguy | just as a heads up its liberty-1 tagging time, let me know if you see things that mean we shouldn't tag | 16:34 |
| *** ijw_ has joined #openstack-nova | 16:35 | |
| johnthetubaguy | mriedem: I guess there are no major bugs for gate, or needing to go onto stable that we should wait for before cutting liberty-1? | 16:35 |
| *** _kiran_ has quit IRC | 16:36 | |
| *** _kiran_ has joined #openstack-nova | 16:36 | |
| sdague | mriedem: what was it? | 16:37 |
| *** beagles_brb is now known as beagles | 16:38 | |
| *** varunarya has joined #openstack-nova | 16:38 | |
| *** jlanoux has quit IRC | 16:38 | |
| *** Flcn has quit IRC | 16:39 | |
| *** arnaud____ has quit IRC | 16:40 | |
| *** Flcn has joined #openstack-nova | 16:41 | |
| openstackgerrit | Zhang Ni proposed openstack/nova-specs: Improve nova attach volume api to support disk-bus https://review.openstack.org/185973 | 16:41 |
| *** fawadk has joined #openstack-nova | 16:42 | |
| *** rotbeard has quit IRC | 16:42 | |
| *** fawadkhaliq has quit IRC | 16:42 | |
| mriedem | sdague: there is a race where instance A is releasing fixed ip X at the same time that instance B is associating with fixed ip X, and before the fixed IP is marked as allocated to instance B is released and disassociated in the DB (tihnking from instance A but at that point it's associated with instance B) | 16:43 |
| *** davideagnello has quit IRC | 16:43 | |
| *** jasondotstar has quit IRC | 16:43 | |
| *** dsanders has joined #openstack-nova | 16:43 | |
| *** kiranr has joined #openstack-nova | 16:43 | |
| mriedem | sdague: and since the dhcp agent callback doesn't have memory of which instance it should be disassociating from, it just disassociates from $whatevs | 16:43 |
| mriedem | and then kablammo | 16:43 |
| sdague | nice | 16:43 |
| *** atuvenie__ has joined #openstack-nova | 16:44 | |
| *** davideagnello has joined #openstack-nova | 16:44 | |
| *** oro has joined #openstack-nova | 16:45 | |
| mriedem | so i guess there could be a couple of fixes....we could keep track of which instance should be associated with an IP when releasing it and if those don't line up we don't disassociate b/c we assume it's changed, | 16:46 |
| mriedem | or we could just re-associate when the new instance B marks the fixed_ip as allocated=True | 16:46 |
| ndipanov | dansmith, so regarding this https://review.openstack.org/#/c/193576/ | 16:46 |
| ndipanov | I just remembered one point | 16:46 |
| *** baoli has quit IRC | 16:46 | |
| *** _kiran_ has quit IRC | 16:47 | |
| *** Mike_D_laptop has joined #openstack-nova | 16:47 | |
| ndipanov | if we do end up creating a migration_context | 16:47 |
| ndipanov | which I think we should and I think it should live on the instance | 16:47 |
| ndipanov | then our queries in the resource tracker look like this: | 16:47 |
| ndipanov | select my migrations | 16:48 |
| ndipanov | for mig in migrations | 16:48 |
| ndipanov | select my instance | 16:48 |
| ndipanov | to get to the needed data | 16:48 |
| ndipanov | now we do this already even though it;s not ideal | 16:48 |
| ndipanov | so we are not breaking anything | 16:48 |
| dansmith | ndipanov: we can optimize that in MigrationList to fill out a lazy property of .instance on the migrations in a single go, right? | 16:49 |
| openstackgerrit | Dmitry Guryanov proposed openstack/nova-specs: virtozzo storage volume driver in libvirt https://review.openstack.org/184295 | 16:49 |
| *** unicell has joined #openstack-nova | 16:49 | |
| *** marzif_ has joined #openstack-nova | 16:50 | |
| *** ctrath has quit IRC | 16:50 | |
| *** baoli has joined #openstack-nova | 16:51 | |
| *** krtaylor has joined #openstack-nova | 16:51 | |
| *** ildikov has quit IRC | 16:51 | |
| *** jasondotstar has joined #openstack-nova | 16:52 | |
| ndipanov | dansmith, yes that was my idea basically | 16:54 |
| ndipanov | we should do it anyway really | 16:54 |
| dansmith | aye | 16:54 |
| *** promulo has joined #openstack-nova | 16:54 | |
| ndipanov | note that some but not all of these isntances are likely already loaded by the previous query | 16:55 |
| ndipanov | so maybe the other way around might acutally be better | 16:55 |
| *** apuimedo has joined #openstack-nova | 16:55 | |
| *** brad[] has quit IRC | 16:55 | |
| danpb | dansmith: am I right in thinking the tenant user end can inject arbitrary key=value data to the guest, via cloud-init / metadata service ? | 16:55 |
| dansmith | danpb: yes | 16:56 |
| *** sudipto has joined #openstack-nova | 16:56 | |
| danpb | ah yeah, the --meta key=val arg to nova boot, populates a /meta_data.json file | 16:56 |
| dansmith | danpb: limited to 255 chars per, and limited by quota | 16:56 |
| *** karimb has quit IRC | 16:56 | |
| mriedem | vishy: are you around for a nova-network question? | 16:58 |
| *** BobBall is now known as BobBall_AWOL | 16:59 | |
| mriedem | damn, meeting | 16:59 |
| mriedem | dansmith: if you have a sec, do you have any idea why we can't mark the fixed ip as allocated=True during associate here? http://git.openstack.org/cgit/openstack/nova/tree/nova/network/manager.py#n905 | 16:59 |
| mriedem | does the vif have to be created first? | 17:00 |
| *** patrickeast has joined #openstack-nova | 17:00 | |
| mriedem | basically, associate, vif and allocated need to be atomic | 17:00 |
| mriedem | and they aren't today | 17:00 |
| mriedem | brb | 17:00 |
| dansmith | mriedem: maybe the async network allocate in compute manager? | 17:00 |
| *** zz_jgrimm is now known as jgrimm | 17:01 | |
| dansmith | oh, no | 17:01 |
| dansmith | we do it right below | 17:01 |
| *** whenry has joined #openstack-nova | 17:01 | |
| *** dontalton has joined #openstack-nova | 17:01 | |
| dansmith | mriedem: so create the vif, and pass it into the associate call to do it atomically? | 17:02 |
| *** electrocucaracha has quit IRC | 17:02 | |
| dansmith | mriedem: does associate need to be atomic with the rest of those things, or just cleanup on failure? | 17:03 |
| *** baoli has quit IRC | 17:03 | |
| *** ctrath has joined #openstack-nova | 17:03 | |
| *** husanu has quit IRC | 17:03 | |
| *** agireud has quit IRC | 17:04 | |
| *** agireud has joined #openstack-nova | 17:05 | |
| *** matrohon has quit IRC | 17:05 | |
| *** husanu has joined #openstack-nova | 17:05 | |
| *** fawadk has quit IRC | 17:06 | |
| *** veena has quit IRC | 17:10 | |
| hemna | sdague, any way to see what was in the /etc/nova/* for a gate-nova-python27 check? I don't get why my patch is failing the py27 test upstream http://logs.openstack.org/69/175569/17/check/gate-nova-python27/520b9ad/console.html | 17:10 |
| hemna | sdague, for this patch https://review.openstack.org/#/c/175569/17 I did touch the compute.filters, but the commands rootwrap is complaining about are in the computer.filters file. | 17:12 |
| sdague | hemna: you are missing mocks for rootwrap iscsi calls | 17:12 |
| sdague | http://logs.openstack.org/69/175569/17/check/gate-nova-python27/520b9ad/console.html#_2015-06-22_23_48_07_428 | 17:12 |
| hemna | hrmm | 17:13 |
| hemna | I can't reproduce this locally | 17:13 |
| hemna | I run tox -epy27 and don't get any complaints | 17:13 |
| sdague | right, because there is an extra check in the gate to make sure you didn't try to run sudo | 17:13 |
| *** electrocucaracha has joined #openstack-nova | 17:14 | |
| hemna | ugh, wish it would say what test it was running that caused that failure | 17:14 |
| sdague | so where new rootwrap calls are being added in your code, you need to mock those out | 17:14 |
| sdague | well, it doesn't know | 17:14 |
| sdague | it just knows sudo attempts were in syslog | 17:15 |
| sdague | but you can see what commands are run | 17:15 |
| sdague | and you know what files you changed | 17:15 |
| sdague | cross referencing shouldn't be too bad | 17:15 |
| hemna | so those are just dumps from syslog it found ? | 17:15 |
| sdague | yep | 17:15 |
| hemna | ah ok | 17:15 |
| sdague | sudo /usr/local/jenkins/slave_scripts/jenkins-sudo-grep.sh post | 17:15 |
| sdague | that's what goes and tries to find those things | 17:15 |
| hemna | ok thanks, I was confused where that was coming from | 17:16 |
| hemna | which probably means I have them in my local syslog | 17:16 |
| *** angdraug has joined #openstack-nova | 17:16 | |
| hemna | just didn't know they were in there. | 17:16 |
| hemna | ok thanks | 17:16 |
| sdague | no prob | 17:16 |
| openstackgerrit | Tracy Jones proposed openstack/nova: Cleanup of compute api reboot method https://review.openstack.org/127427 | 17:16 |
| *** mleroy_ has quit IRC | 17:18 | |
| *** kiranr has quit IRC | 17:19 | |
| *** dboik_ has quit IRC | 17:20 | |
| *** Flcn has quit IRC | 17:23 | |
| *** dboik has joined #openstack-nova | 17:24 | |
| *** mestery has joined #openstack-nova | 17:25 | |
| *** harlowja has joined #openstack-nova | 17:26 | |
| *** ildikov has joined #openstack-nova | 17:27 | |
| *** tjones1 has joined #openstack-nova | 17:28 | |
| openstackgerrit | Tracy Jones proposed openstack/nova: Cleanup of compute api reboot method https://review.openstack.org/127427 | 17:28 |
| *** baoli has joined #openstack-nova | 17:29 | |
| *** VW_ has quit IRC | 17:30 | |
| mriedem | dansmith: i think just cleanup on failure | 17:31 |
| *** rwsu has joined #openstack-nova | 17:31 | |
| mriedem | dansmith: i was thinking the same, create the vif and pass that into associate | 17:31 |
| *** electrocucaracha has quit IRC | 17:31 | |
| mriedem | and if it's there, mark the fixed ip as allocated=True when we associate | 17:31 |
| mriedem | dansmith: although that would be an api change to the fixed ip object so we couldn't backport it as far as i know | 17:32 |
| *** gyee_ has quit IRC | 17:33 | |
| *** rwsu has quit IRC | 17:33 | |
| mriedem | we could definitely reduce the scope of the race by getting the vif up front before calling associate | 17:34 |
| mriedem | but w/o a lock or a change to the db api it's not a complete fix of the race | 17:34 |
| *** gyee has joined #openstack-nova | 17:36 | |
| dansmith | mriedem: yeah.. if it doesn't need to be atomic, just clean-up-able, we can just do a better job there of catching failure and undoing the association, right? | 17:37 |
| dansmith | maybe a helper method and we try..except: fip.disassociate() ? | 17:38 |
| *** rwsu has joined #openstack-nova | 17:39 | |
| *** ZZelle has joined #openstack-nova | 17:39 | |
| *** rwsu has quit IRC | 17:42 | |
| *** varunarya has quit IRC | 17:42 | |
| *** rwsu has joined #openstack-nova | 17:42 | |
| *** tonytan4ever has joined #openstack-nova | 17:43 | |
| mriedem | dansmith: there isn't a failure | 17:44 |
| mriedem | that's the thing | 17:44 |
| *** marun has joined #openstack-nova | 17:44 | |
| mriedem | the failure comes later when we try to associate a floating IP and there is no nw_info in the instance_info_cache | 17:44 |
| mriedem | b/c get_instance_nw_info can't build the nw info model since there is no fixed IP association | 17:44 |
| dansmith | okay I'm confused | 17:45 |
| mriedem | we could definitely fail earlier in this process if by the time we leave allocate_for_instance, there is no nw_info to return | 17:45 |
| openstackgerrit | Qin Zhao proposed openstack/nova: Use flavor object in compute manager https://review.openstack.org/194761 | 17:45 |
| mriedem | but i just wanted to fix the race | 17:45 |
| mriedem | dansmith: this should show the failure steps https://bugs.launchpad.net/nova/+bug/1249065/comments/45 | 17:46 |
| openstack | Launchpad bug 1249065 in OpenStack Compute (nova) "Nova throws 400 when attempting to add floating ip (instance.info_cache.network_info is empty)" [Medium,Triaged] - Assigned to Matt Riedemann (mriedem) | 17:46 |
| mriedem | to fix the race, associated/vif/allocated need to be set at the same time in the db api | 17:46 |
| mriedem | or a lock....but i'd like to avoid a lock | 17:46 |
| *** irenab has joined #openstack-nova | 17:47 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Use flavor object in compute manager https://review.openstack.org/194761 | 17:47 |
| mriedem | anywho, i think i have the steps to move forward with 2 patches, first patch moves vif allocation before associate, that narrows the race but doesn't fix it - but could be backported to stable/kilo, | 17:47 |
| dansmith | mriedem: hmm, wow, that takes some thinkin' | 17:47 |
| mriedem | 2nd patch updates the associate rpc api/db api to pass in the vif | 17:48 |
| mriedem | if vif: allocated=True | 17:48 |
| mriedem | during association | 17:48 |
| mriedem | yeah, the dhcpagent callback is the wild card | 17:48 |
| mriedem | i loves me some externally managed callbacks | 17:48 |
| *** dramakri has joined #openstack-nova | 17:49 | |
| dansmith | mriedem: so is it that this IP was used by one instance, then another in quick succession and the dhcp_agent callback from the first breaks the second? | 17:49 |
| mriedem | yup | 17:49 |
| dansmith | is the bug then that we allow that to happen? like, the first instance's dhcp_agent shouldn't break the allocation of the second, right? | 17:49 |
| *** armax has joined #openstack-nova | 17:50 | |
| *** tjones1 has quit IRC | 17:50 | |
| mriedem | like, http://logs.openstack.org/42/175742/10/check/check-tempest-dsvm-nova-v21-full/2850b02/logs/screen-n-net.txt.gz#_2015-06-23_05_10_55_459 | 17:50 |
| mriedem | when we get to http://git.openstack.org/cgit/openstack/nova/tree/nova/network/manager.py#n1114 | 17:51 |
| mriedem | if it's not allocated and it's not leased, don't disassociate | 17:51 |
| dansmith | well, | 17:51 |
| mriedem | i think the reason i was avoiding that was because those warnings show up a ton in non failure runs | 17:51 |
| dansmith | more like, I feel like we should know what instance is trying to do the release | 17:52 |
| dansmith | that release code seems pretty fragile and assume-y | 17:52 |
| mriedem | dansmith: yeah, and we don't have that info | 17:52 |
| mriedem | yes, very much so | 17:52 |
| dansmith | er, presumptuous | 17:52 |
| mriedem | yeah, it's a real dick :) | 17:52 |
| dansmith | haha | 17:52 |
| mriedem | http://goo.gl/oh4pY8 | 17:52 |
| mriedem | 84% of the time that shows up, it's not in a failing job | 17:53 |
| mriedem | and it shows up a ton | 17:53 |
| dansmith | that doesn't mean it's right | 17:53 |
| mriedem | i know, i'm sure there is some other legacy whack a mole check in n-net somewhere to handle it | 17:53 |
| dansmith | like, it could be causing other issues depending on when it's failing I'd think | 17:53 |
| mriedem | hrm, i wonder how much this race contributes to our old buddy the ssh timeout bug | 17:54 |
| dansmith | right | 17:54 |
| dansmith | that's what I was thinking | 17:54 |
| *** toure has quit IRC | 17:54 | |
| dansmith | although that happens on neutron too | 17:54 |
| mriedem | yeah | 17:54 |
| dansmith | but it's also hard to define, so maybe this is related to the nova-network side of the failure | 17:54 |
| *** toure has joined #openstack-nova | 17:54 | |
| *** fawadkhaliq has joined #openstack-nova | 17:54 | |
| *** toure has quit IRC | 17:55 | |
| mriedem | right, so another option i was thinking of was tracking in the network manager a map of associated fixed_ip/instance relationships before and after a lease/release, so when we hit http://git.openstack.org/cgit/openstack/nova/tree/nova/network/manager.py#n1114 - if it's not associated with what we thought it should be in the cache, we don't disassociate | 17:55 |
| mriedem | but i feel like that is racy also | 17:56 |
| dansmith | si | 17:56 |
| dansmith | so, just looking at how this is plumbed makes me want to stab something | 17:56 |
| *** Nikolay_St has quit IRC | 17:56 | |
| *** toure has joined #openstack-nova | 17:56 | |
| mriedem | i'm pretty confident that making the associated/allocated thing atomic would fix this | 17:56 |
| dansmith | I think it would fix this race, | 17:57 |
| dansmith | but it doesn't fix the fragility of what is going on here | 17:57 |
| dansmith | let me look at some things, just a sec | 17:57 |
| mriedem | oh yeah, btw, this would also have to be fixed for the VlanManager :( | 17:58 |
| mriedem | http://git.openstack.org/cgit/openstack/nova/tree/nova/network/manager.py#n1939 | 17:58 |
| mriedem | since it's allocate_fixed_ip method follows the same pattern | 17:59 |
| mriedem | it's just that we test FlatDHCPManager in the gate | 17:59 |
| dansmith | so, what I was thinking | 17:59 |
| *** busterswt has quit IRC | 17:59 | |
| *** mpaolino has joined #openstack-nova | 17:59 | |
| dansmith | is that we write out information that tells dhcpbridge what instance currently owns an IP, so that we can pass in the instance_uuid when we make that call | 17:59 |
| *** mtanino has quit IRC | 18:00 | |
| dansmith | or, | 18:00 |
| openstackgerrit | Joe Gordon proposed openstack/nova-specs: trivial: Change file permissions for spec https://review.openstack.org/194559 | 18:00 |
| *** mestery has quit IRC | 18:00 | |
| *** Mike_D_laptop1 has joined #openstack-nova | 18:00 | |
| *** mestery has joined #openstack-nova | 18:01 | |
| *** Mike_D_laptop has quit IRC | 18:01 | |
| *** mwagner_afk has quit IRC | 18:02 | |
| dansmith | mriedem: we have the mac address of the thing doing the delete | 18:02 |
| *** otter768 has joined #openstack-nova | 18:02 | |
| dansmith | mriedem: we should pass that over rpc and not do the thing if they don't match, no? | 18:02 |
| *** ociuhandu has quit IRC | 18:03 | |
| dansmith | I can cook up that change if you want | 18:03 |
| mriedem | no i got this one :) | 18:03 |
| mriedem | dibs | 18:03 |
| dansmith | uh! | 18:03 |
| mriedem | let me check out the mac thing | 18:03 |
| dansmith | doesn't the prescriber of the fix get automatic dibs? | 18:04 |
| *** ericksonfgds is now known as ericksonsantos | 18:04 | |
| *** electrocucaracha has joined #openstack-nova | 18:04 | |
| mriedem | let me consult the playground rules handbook quick | 18:04 |
| *** iamjarvo has joined #openstack-nova | 18:05 | |
| *** armax has quit IRC | 18:05 | |
| *** achanda_ has joined #openstack-nova | 18:06 | |
| mriedem | dansmith: i don't think that will work | 18:06 |
| mriedem | the mac is on the vif | 18:06 |
| mriedem | and when we hit this: http://git.openstack.org/cgit/openstack/nova/tree/nova/network/manager.py#n1114 | 18:06 |
| mriedem | the vif isn't associated with the fixed IP yet | 18:06 |
| *** nic has quit IRC | 18:06 | |
| *** VW_ has joined #openstack-nova | 18:06 | |
| mriedem | that happens here http://git.openstack.org/cgit/openstack/nova/tree/nova/network/manager.py#n920 | 18:07 |
| dansmith | mriedem: you look up the vif based on the mac, which gives you the instance_uuid, which you can compare to the association, right? | 18:07 |
| *** jasondotstar has quit IRC | 18:07 | |
| *** otter768 has quit IRC | 18:07 | |
| mriedem | dansmith: ah, gotcha | 18:07 |
| mriedem | sounds like backdoor action to me | 18:07 |
| mriedem | i like it | 18:07 |
| dansmith | heh | 18:08 |
| jogo | jaypipes: your thoughts on https://review.openstack.org/#/c/191914/ would be appreciated | 18:08 |
| dansmith | I totally get a co-authored-by on this if no dibs | 18:08 |
| *** achanda has quit IRC | 18:08 | |
| dansmith | jogo: when is (or was) your last day? | 18:09 |
| mriedem | dansmith: so, this is still an rpc api change to release_fixed_ip right - which means no backport, which is fine, but i'm looking for "something" to backport, which i think could be at least re-arranging the order of the vif create and the fixed_ip.associate call | 18:09 |
| mriedem | dansmith: yeah i was going to co-author | 18:09 |
| dansmith | mriedem: yeah, I think you change the ordering for the backport to narrow the window and then fix it for realz with the rpc change | 18:09 |
| mriedem | cool, just checking | 18:09 |
| mriedem | btw, i don't think jogo is actually quitting, | 18:09 |
| mriedem | he just said he is to be left alone | 18:10 |
| dansmith | mriedem: we *can* backport things like this if we need to | 18:10 |
| *** mestery has quit IRC | 18:10 | |
| dansmith | but it doesn't sound like this is serious enough, especially if you narrow it significantly with the easy part | 18:10 |
| mriedem | yeah | 18:10 |
| mriedem | well, that's the hope anyway | 18:11 |
| mriedem | b/c the db api call or vif create is taking just long enough to fudge it up | 18:11 |
| dansmith | yeah | 18:11 |
| jogo | dansmith: friday | 18:11 |
| dansmith | the proper fix will also make this not super-effing-scary in terms of one instance un-associating another's fixed | 18:11 |
| jogo | mriedem: hehe | 18:11 |
| dansmith | jogo: this coming friday? | 18:11 |
| *** agireud has quit IRC | 18:11 | |
| jogo | dansmith: the 26th | 18:12 |
| dansmith | okay, so we still have a few days to talk you out of it, right? | 18:12 |
| mriedem | we should have a going away party | 18:12 |
| jogo | dansmith: yup | 18:12 |
| jogo | mriedem: yup | 18:12 |
| dansmith | jogo: don't do it. | 18:12 |
| *** moshele_ has joined #openstack-nova | 18:13 | |
| *** armax has joined #openstack-nova | 18:13 | |
| *** jasondotstar has joined #openstack-nova | 18:14 | |
| *** mpaolino has quit IRC | 18:14 | |
| jogo | dansmith: looking over the specs there are just so many I don't want to deal with | 18:15 |
| jogo | such as https://review.openstack.org/#/c/194453/2 | 18:15 |
| jogo | and that isn't even a bad one | 18:15 |
| mriedem | bug jogo, the gate is relatively benign now, stable dep hell isn't happening so much anymore | 18:16 |
| mriedem | jogo: there is just so much to live for now | 18:16 |
| jogo | mriedem: haha | 18:16 |
| mriedem | we might have a fix for the ssh timeout n-net bug finally | 18:16 |
| mriedem | come on | 18:16 |
| jogo | mriedem: that reminds me have to revsit what is breaking in multinode | 18:16 |
| jogo | mriedem: btw http://imgh.us/nova.svg | 18:16 |
| *** fawadkhaliq has quit IRC | 18:16 | |
| *** artom has joined #openstack-nova | 18:16 | |
| mtreinish | mriedem: I don't believe you, that bug will never be fixed | 18:16 |
| jogo | preliminary diagram | 18:16 |
| mriedem | ascii, please | 18:16 |
| *** Marga_ has quit IRC | 18:17 | |
| *** achanda_ has quit IRC | 18:17 | |
| *** amotoki has quit IRC | 18:17 | |
| *** agireud has joined #openstack-nova | 18:17 | |
| mtreinish | jogo: I think you need a legend | 18:17 |
| mtreinish | unless you know what you're looking at the different line types don't mean anything | 18:17 |
| *** achanda has joined #openstack-nova | 18:17 | |
| *** armax has quit IRC | 18:18 | |
| jogo | mtreinish: yup, I do | 18:18 |
| *** danpb has quit IRC | 18:19 | |
| jogo | was trying to replace http://docs.openstack.org/developer/nova/architecture.html | 18:19 |
| *** artom has quit IRC | 18:21 | |
| *** tjones1 has joined #openstack-nova | 18:22 | |
| jogo | haha neutron unit tests are broken | 18:23 |
| *** artom has joined #openstack-nova | 18:24 | |
| *** Flcn has joined #openstack-nova | 18:24 | |
| openstackgerrit | Dan Smith proposed openstack/nova-specs: Add keep-jogo spec https://review.openstack.org/194774 | 18:26 |
| dansmith | request -2s and -1s please ^^ | 18:26 |
| dansmith | er, wait, +2s and +1s I guess :D | 18:26 |
| mriedem | ha | 18:27 |
| *** lpetrut1 has quit IRC | 18:27 | |
| mriedem | previously-next-employer? | 18:28 |
| dansmith | yeah, assuming he changes his mind, they'd be previously-next | 18:28 |
| jogo | so this is fun https://bugs.launchpad.net/nova/+bug/1462305 | 18:32 |
| openstack | Launchpad bug 1462305 in OpenStack Compute (nova) "multi-node test causes nova-compute to lockup" [High,In progress] - Assigned to Joe Gordon (jogo) | 18:32 |
| jogo | the hung nova-compute won't respond to SIGUSR1 | 18:32 |
| jogo | so no guru meditation for me | 18:32 |
| jogo | dansmith: nice spec | 18:33 |
| dansmith | \o/ | 18:33 |
| *** annashen has quit IRC | 18:33 | |
| *** ijw_ has quit IRC | 18:35 | |
| jogo | any ideas on how to further debug https://bugs.launchpad.net/nova/+bug/1462305 ? | 18:36 |
| openstack | Launchpad bug 1462305 in OpenStack Compute (nova) "multi-node test causes nova-compute to lockup" [High,In progress] - Assigned to Joe Gordon (jogo) | 18:36 |
| *** markmcclain has quit IRC | 18:36 | |
| *** burt has quit IRC | 18:38 | |
| *** burt has joined #openstack-nova | 18:38 | |
| dansmith | jogo: hung such that it won't do SIGUSR1 means probably blocked in C-land right? | 18:39 |
| jogo | dansmith: or something funny with eventlet? | 18:39 |
| jogo | dansmith: but that makes sense to me | 18:39 |
| dansmith | could be either yawh | 18:39 |
| dansmith | er, yeah | 18:39 |
| dansmith | so a gdb-attach stack trace maybe if it's available | 18:40 |
| dansmith | the gdb stack trace would tell you if it's handling the signal and blocking on something in eventlet too I think | 18:40 |
| jogo | harlowja: have an eventlet question for you if your around | 18:41 |
| harlowja | jogo uh oh, whats up | 18:42 |
| * harlowja may or may not know, ha | 18:42 | |
| jogo | harlowja: how does eventlet work with signal.signal | 18:42 |
| jogo | https://docs.python.org/2/library/signal.html | 18:42 |
| *** aix has quit IRC | 18:42 | |
| *** burt has quit IRC | 18:42 | |
| jogo | harlowja: we have a hung nova-compute that isn't responding to SIGUSR1 | 18:42 |
| *** marzif_ has quit IRC | 18:43 | |
| *** burt has joined #openstack-nova | 18:43 | |
| harlowja | hmmm, wonder if something is absorbing it | 18:44 |
| jogo | harlowja: when nova-compute is running normally SIGUSR1 works | 18:44 |
| *** lyarwood has quit IRC | 18:44 | |
| jogo | harlowja: just when nova-compute is hung, it breaks | 18:44 |
| *** lyarwood has joined #openstack-nova | 18:44 | |
| jogo | harlowja: example http://logs.openstack.org/40/194540/1/check/check-tempest-dsvm-multinode-full/1c35401/logs/screen-n-cpu.txt.gz | 18:44 |
| harlowja | any idea where hung in that log? | 18:45 |
| jogo | harlowja: iptables (last line) | 18:45 |
| harlowja | k | 18:45 |
| jogo | harlowja: no output after that | 18:46 |
| openstackgerrit | Ed Leafe proposed openstack/nova: Make NoValidHost exceptions clearer https://review.openstack.org/194780 | 18:46 |
| harlowja | hmmm, thinking | 18:47 |
| jogo | harlowja: thanks | 18:47 |
| harlowja | np | 18:47 |
| jogo | if this was just a deadlock in how we use ... locks, then I would expect the sigusr1 (GMR) stuff to work | 18:48 |
| harlowja | https://github.com/openstack/nova/blob/master/nova/network/linux_net.py#L446 that code i guess? | 18:48 |
| jogo | but maybe some eventlet thing is preventing that | 18:48 |
| jogo | harlowja: roughly yeah. only see this in multihost/multinode | 18:48 |
| harlowja | k | 18:48 |
| *** moshele_ has quit IRC | 18:49 | |
| jogo | dansmith: I could do gdb, would have to get infra to help though | 18:50 |
| dansmith | jogo: yeah | 18:50 |
| *** belmoreira has joined #openstack-nova | 18:51 | |
| jogo | dansmith: since running out of ideas I think that may be the next step | 18:51 |
| *** oro has quit IRC | 18:51 | |
| *** rwsu has quit IRC | 18:51 | |
| harlowja | ya, i'd be interested in knowing where it is at | 18:51 |
| *** irenab has quit IRC | 18:52 | |
| *** nihilifer has joined #openstack-nova | 18:52 | |
| *** rwsu has joined #openstack-nova | 18:53 | |
| *** electrocucaracha has quit IRC | 18:53 | |
| harlowja | i've seen similar kind of lockups when things aren't monkey patched, but this doesn't seem to be that | 18:54 |
| harlowja | *or when certain things are monkey patched, and certain things aren't | 18:54 |
| *** baoli has quit IRC | 18:54 | |
| harlowja | any chance that could be part of it? | 18:54 |
| jogo | harlowja: the lockup part probably not. but GMR not working ... let me check | 18:55 |
| *** baoli has joined #openstack-nova | 18:56 | |
| harlowja | i do see https://github.com/eventlet/eventlet/blob/master/eventlet/hubs/hub.py#L8 but thats probably ok (periodic alarms) | 18:56 |
| jogo | harlowja: we monkeypatch before calling gmr | 18:56 |
| harlowja | monkey patch all the things? or just some of them? | 18:56 |
| jogo | harlowja: see nova/cmd/compute.py | 18:57 |
| harlowja | k | 18:58 |
| harlowja | hmmm, i didn't think the 'utils.monkey_patch()' was eventlet monkey patching | 18:58 |
| harlowja | although i hope https://github.com/openstack/nova/blob/master/nova/cmd/__init__.py does whats needed | 18:59 |
| jogo | oh right | 18:59 |
| *** nic has joined #openstack-nova | 18:59 | |
| harlowja | although 'eventlet.monkey_patch(os=False, thread=False)' is bad, but ya, oh well | 19:00 |
| thomasem | So, looks like novaclient doesn't honor bypass_url anymore? | 19:00 |
| jogo | harlowja: anyway stepping out to lunch, if you have any ideas they would be appreciated. But for now i think attaching gdb is the best bet | 19:00 |
| harlowja | jogo agreed, gdb would be really useful | 19:00 |
| *** electrocucaracha has joined #openstack-nova | 19:01 | |
| harlowja | also figuring out which eventlet hub thats using might be useful to | 19:02 |
| harlowja | probably should dump that out somewhere | 19:02 |
| harlowja | *in the future | 19:02 |
| *** mgarza_ has quit IRC | 19:04 | |
| *** Nikolay_St has joined #openstack-nova | 19:05 | |
| *** rwsu has quit IRC | 19:05 | |
| *** Longgeek_ has quit IRC | 19:07 | |
| melwitt | thomasem: it should, if it's not, it's a bug. what version are you using? | 19:08 |
| thomasem | melwitt: looking into it. 2.26.0, I see a fix for this problem going in to 2.25.0, so it may be some other environmental issue | 19:09 |
| *** electrocucaracha has quit IRC | 19:10 | |
| *** busterswt has joined #openstack-nova | 19:11 | |
| *** whenry has quit IRC | 19:11 | |
| *** markmcclain has joined #openstack-nova | 19:11 | |
| *** dramakri has left #openstack-nova | 19:11 | |
| melwitt | thomasem: are you using --bypass-url to specify the bypass url? if that doesn't work, it would help us if you open a bug against novaclient about it | 19:12 |
| *** LukeHinds has quit IRC | 19:12 | |
| thomasem | melwitt: false alarm, I think :) | 19:12 |
| *** Mike_D_laptop1 has quit IRC | 19:12 | |
| thomasem | melwitt: sorry for that | 19:12 |
| *** markmcclain has quit IRC | 19:12 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova-specs: Add rootwrap-daemon-mode blueprint https://review.openstack.org/105404 | 19:12 |
| thomasem | melwitt: If I do discover it's a problem, I'll open a bug and probably submit a patch for it. | 19:12 |
| thomasem | melwitt: Thanks for your time on it | 19:12 |
| *** ctrath has left #openstack-nova | 19:13 | |
| *** dramakri has joined #openstack-nova | 19:13 | |
| *** markmcclain has joined #openstack-nova | 19:13 | |
| melwitt | thomasem: okay, I will be sad if that's broken again. :) sure, sounds good | 19:13 |
| *** ayoung has joined #openstack-nova | 19:13 | |
| thomasem | LOL, yeah, right? | 19:13 |
| *** rwsu has joined #openstack-nova | 19:13 | |
| *** ctrath has joined #openstack-nova | 19:13 | |
| *** atuvenie__ has quit IRC | 19:13 | |
| *** ctrath has left #openstack-nova | 19:15 | |
| *** mgarza_ has joined #openstack-nova | 19:15 | |
| *** sudipto has quit IRC | 19:16 | |
| thomasem | melwitt: confirmed false alarm, just saw the requests go through fine. | 19:16 |
| thomasem | Sleep easy | 19:16 |
| melwitt | thomasem: haha, yay | 19:16 |
| thomasem | huzzah | 19:17 |
| mriedem | i thought mox was strict about the order of calls? | 19:17 |
| mriedem | i.e. if A is called before B but you mox then out in the opposite order it would fail? | 19:17 |
| *** scheuran has joined #openstack-nova | 19:17 | |
| harlowja | jogo http://bugs.python.org/issue13697 may be releated (not sure); getting that gdb backtrace would be great :) | 19:18 |
| *** busterswt has quit IRC | 19:18 | |
| harlowja | if ^ is releavant, lifeless might have some idears | 19:18 |
| openstackgerrit | Qin Zhao proposed openstack/nova: Make compute rpc api to send instance object https://review.openstack.org/194791 | 19:21 |
| *** openstackgerrit has quit IRC | 19:21 | |
| artom | mriedem, thanks for your review of the all tenants floating IPs spec. | 19:22 |
| mriedem | artom: np | 19:22 |
| *** openstackgerrit has joined #openstack-nova | 19:22 | |
| artom | I'm wondering what you meant by a change in policy.json having a security impact... | 19:22 |
| artom | The spec template doesn't really mention policy.json in the security section. | 19:22 |
| *** yamahata has joined #openstack-nova | 19:23 | |
| mriedem | artom: well, maybe it's just a deployment impact, but i'd think that changes to the default policy.json are security related | 19:23 |
| artom | Should I just mention "it's an admin API by default, it's proabably a bad idea for operators to open this up to everyone" ? | 19:23 |
| *** yamahata has quit IRC | 19:24 | |
| mriedem | artom: doesn't there have to be an entry in the default policy.json and then the floating_ips extension has to check for that policy setting when all_tenants is used? | 19:24 |
| mriedem | like in fping | 19:24 |
| *** yamahata has joined #openstack-nova | 19:24 | |
| *** ctrath has joined #openstack-nova | 19:24 | |
| mriedem | artom that's why i added ken'ichi | 19:25 |
| mriedem | for the api questions | 19:25 |
| mriedem | might want to ask sdague about the policy.json question | 19:25 |
| *** achanda has quit IRC | 19:26 | |
| artom | mriedem, that pinged him. Do I need to do it twice more, like for Cthulhu? | 19:26 |
| artom | ;) | 19:26 |
| *** husanu has quit IRC | 19:26 | |
| mriedem | no, more than once makes him swallow your soul in the water deep | 19:26 |
| mriedem | *watery | 19:26 |
| artom | Not much left to swallow... | 19:27 |
| *** lpetrut has joined #openstack-nova | 19:27 | |
| *** husanu has joined #openstack-nova | 19:28 | |
| *** lyarwood has quit IRC | 19:29 | |
| *** lyarwood has joined #openstack-nova | 19:29 | |
| *** fifieldt_ has joined #openstack-nova | 19:31 | |
| artom | mriedem, next point! The ceilometer thing was reverted out, no? https://review.openstack.org/#/c/99212/ | 19:32 |
| artom | So it shouldn't be polling the API anymore (anyways the API hasn't existed for like a year) | 19:32 |
| mriedem | artom: i assumed you were going to be adding that back into ceilometer | 19:33 |
| mriedem | hence the need for the api in onva | 19:33 |
| mriedem | *nova | 19:33 |
| artom | mriedem, nope, it's users who are asking it. | 19:33 |
| artom | Apparently some admins want to do it, since the client advertised that we could | 19:33 |
| *** fifieldt has quit IRC | 19:34 | |
| *** annashen has joined #openstack-nova | 19:34 | |
| *** belmoreira has quit IRC | 19:35 | |
| *** Mike_D_laptop has joined #openstack-nova | 19:36 | |
| bauzas | mriedem: heard about nova.tests.unit.objects.test_instance.TestRemoteInstanceObject.test_save_objectfield_missing_instance_row failure ? | 19:36 |
| mriedem | bauzas: heard about it? | 19:36 |
| mriedem | no | 19:36 |
| bauzas | mriedem: tried to see logstash and there are some failures | 19:36 |
| bauzas | mriedem: http://logstash.openstack.org/#eyJzZWFyY2giOiJidWlsZF9uYW1lOlwiZ2F0ZS1ub3ZhLXB5dGhvbjI3XCIgQU5EIG1lc3NhZ2U6XCJ0ZXN0dG9vbHMubWF0Y2hlcnMuX2ltcGwuTWlzbWF0Y2hFcnJvclwiIEFORCBidWlsZF9zdGF0dXM6XCJGQUlMVVJFXCIiLCJmaWVsZHMiOltdLCJvZmZzZXQiOjAsInRpbWVmcmFtZSI6IjYwNDgwMCIsImdyYXBobW9kZSI6ImNvdW50IiwidGltZSI6eyJ1c2VyX2ludGVydmFsIjowfSwic3RhbXAiOjE0MzUwODgyMjQ1NjN9 | 19:37 |
| *** Longgeek has joined #openstack-nova | 19:38 | |
| bauzas | mriedem: so was wondering if you were aware of it b/c http://status.openstack.org/elastic-recheck/ doesn't provide it | 19:38 |
| *** dprince has quit IRC | 19:38 | |
| bauzas | mriedem: a good opportunity for merging a patch for e-r ? :) | 19:39 |
| mriedem | bauzas: that query is really generic | 19:39 |
| mriedem | any mismatch failure in nova unit tests would hit that | 19:39 |
| *** Marga_ has joined #openstack-nova | 19:39 | |
| *** iamjarvo has quit IRC | 19:39 | |
| *** electrocucaracha has joined #openstack-nova | 19:39 | |
| *** annashen has quit IRC | 19:40 | |
| mriedem | i first need to figure out why mox isn't verifying the order of calls on moxed out calls | 19:40 |
| bauzas | mriedem: build_name:"gate-nova-python27" AND message:"testtools.matchers._impl.MismatchError: <bound method Instance.save of Instance" AND build_status:"FAILURE" would be better | 19:40 |
| *** thorst has quit IRC | 19:40 | |
| *** tonytan4ever has quit IRC | 19:40 | |
| bauzas | mriedem: and you're right, using this new message shows that the failure is really new | 19:41 |
| bauzas | http://logstash.openstack.org/#eyJzZWFyY2giOiJidWlsZF9uYW1lOlwiZ2F0ZS1ub3ZhLXB5dGhvbjI3XCIgQU5EIG1lc3NhZ2U6XCJ0ZXN0dG9vbHMubWF0Y2hlcnMuX2ltcGwuTWlzbWF0Y2hFcnJvcjogJmx0O2JvdW5kIG1ldGhvZCBJbnN0YW5jZS5zYXZlIG9mIEluc3RhbmNlXCIgQU5EIGJ1aWxkX3N0YXR1czpcIkZBSUxVUkVcIiIsImZpZWxkcyI6W10sIm9mZnNldCI6MCwidGltZWZyYW1lIjoiNjA0ODAwIiwiZ3JhcGhtb2RlIjoiY291bnQiLCJ0aW1lIjp7InVzZXJfaW50ZXJ2YWwiOjB9LCJzdGFtcCI6MTQzNTA4ODM5OTU4Mn0= | 19:41 |
| *** rwsu has quit IRC | 19:41 | |
| *** tonytan4ever has joined #openstack-nova | 19:41 | |
| mriedem | bauzas: is there a bug reported? | 19:41 |
| bauzas | mriedem: none I found, so I'm gonna create it and match it for the e-r query | 19:41 |
| mriedem | http://goo.gl/LjaHFH | 19:41 |
| mriedem | yeah, showed up in the last 24 hours | 19:42 |
| mriedem | so what did we merge that touches that code? | 19:42 |
| bauzas | mriedem: lemme see | 19:42 |
| bauzas | mriedem: before doing all the things, I wanted to see if you were aware of that | 19:42 |
| bauzas | mriedem: now that it seems it's a new failure, I'll look at it | 19:42 |
| *** aix has joined #openstack-nova | 19:43 | |
| *** sambol has joined #openstack-nova | 19:43 | |
| *** achanda has joined #openstack-nova | 19:43 | |
| *** tjones1 has quit IRC | 19:43 | |
| openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Support --all-tenants in floatingips-list https://review.openstack.org/171389 | 19:44 |
| *** Mike_D_laptop has quit IRC | 19:44 | |
| *** rwsu has joined #openstack-nova | 19:45 | |
| *** tonytan4ever has quit IRC | 19:45 | |
| *** rdo has quit IRC | 19:45 | |
| *** tonytan4ever has joined #openstack-nova | 19:46 | |
| jaypipes | anybody got the priorities etherpad or wiki page handy?\ | 19:46 |
| jaypipes | bauzas: ^? | 19:46 |
| mriedem | jaypipes: https://etherpad.openstack.org/p/liberty-nova-priorities-tracking | 19:47 |
| jaypipes | danke | 19:47 |
| *** rdo has joined #openstack-nova | 19:47 | |
| bauzas | mriedem: FYI, https://bugs.launchpad.net/nova/+bug/1468087 | 19:48 |
| openstack | Launchpad bug 1468087 in OpenStack Compute (nova) "py27 job failing on test_save_objectfield_missing_instance_row" [High,Confirmed] | 19:48 |
| bauzas | mriedem: creating the e-r query, and trying to bisect the bug | 19:49 |
| lifeless | harlowja: heh, that was a fun investigation :) | 19:50 |
| harlowja | lifeless i bet :) | 19:50 |
| lifeless | jogo: ^ whats the context w.r.t. RLock ? | 19:51 |
| *** tonytan4ever has quit IRC | 19:51 | |
| *** rwsu has quit IRC | 19:51 | |
| lifeless | oh I see, eventlet using the python RLock will naturally suffer the same race | 19:51 |
| harlowja | jogo lifeless https://bugs.launchpad.net/nova/+bug/1462305 | 19:52 |
| openstack | Launchpad bug 1462305 in OpenStack Compute (nova) "multi-node test causes nova-compute to lockup" [High,In progress] - Assigned to Joe Gordon (jogo) | 19:52 |
| harlowja | http://logs.openstack.org/40/194540/1/check/check-tempest-dsvm-multinode-full/1c35401/logs/screen-n-cpu.txt.gz | 19:52 |
| harlowja | ^ dead in water (but jogo hopefully can get a gdb trace from infra?) | 19:52 |
| harlowja | so i was thinking/looking for something that might be releated | 19:52 |
| harlowja | and stumbled into that bug | 19:52 |
| *** tonytan4ever has joined #openstack-nova | 19:52 | |
| *** sdake has quit IRC | 19:54 | |
| dims | jogo: around? can you please peek at this rootwrap spec when you get a chance - https://review.openstack.org/#/c/105404/ | 19:54 |
| dims | jogo: implementation is ready as well (https://review.openstack.org/#/c/180695/) but that can wait | 19:55 |
| *** Marga_ has quit IRC | 19:56 | |
| *** brad[] has joined #openstack-nova | 19:56 | |
| *** Marga_ has joined #openstack-nova | 19:57 | |
| *** tjones1 has joined #openstack-nova | 19:58 | |
| jogo | dims: looking | 19:59 |
| *** gholler has joined #openstack-nova | 20:03 | |
| *** otter768 has joined #openstack-nova | 20:03 | |
| *** annegentle has joined #openstack-nova | 20:04 | |
| *** thangp has quit IRC | 20:05 | |
| *** achanda has quit IRC | 20:06 | |
| *** greghaynes has quit IRC | 20:08 | |
| *** gholler has quit IRC | 20:09 | |
| *** otter768 has quit IRC | 20:09 | |
| *** gholler has joined #openstack-nova | 20:09 | |
| *** ndipanov has quit IRC | 20:10 | |
| *** romainh has joined #openstack-nova | 20:10 | |
| jogo | dims: needs one quick respin then +2 from me | 20:11 |
| *** thorst has joined #openstack-nova | 20:12 | |
| *** Marga_ has quit IRC | 20:13 | |
| *** Flcn has quit IRC | 20:14 | |
| *** Marga_ has joined #openstack-nova | 20:14 | |
| *** Marga_ has quit IRC | 20:15 | |
| *** Marga_ has joined #openstack-nova | 20:15 | |
| *** gszasz has quit IRC | 20:15 | |
| *** prashup has quit IRC | 20:15 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add to_dict() to ImageMeta and ImageMetaProps objects https://review.openstack.org/186474 | 20:16 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add RequestSpec methods for primitiving into dicts https://review.openstack.org/187197 | 20:16 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Use RequestSpec object in HostManager https://review.openstack.org/191251 | 20:16 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Use the same pci_requests field for all filters and HostManager https://review.openstack.org/191250 | 20:16 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Remove unnecessary method in FilterScheduler https://review.openstack.org/188485 | 20:16 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Consider that all scheduler calls are IO Ops https://review.openstack.org/188486 | 20:16 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Create RequestSpec object https://review.openstack.org/145528 | 20:16 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Use RequestSpec object in the FilterScheduler https://review.openstack.org/188487 | 20:16 |
| *** tonytan4ever has quit IRC | 20:16 | |
| *** annegentle has quit IRC | 20:18 | |
| *** annegentle has joined #openstack-nova | 20:18 | |
| bauzas | argh, review.o.o is still flakey | 20:19 |
| *** Sukhdev has joined #openstack-nova | 20:20 | |
| *** iamjarvo has joined #openstack-nova | 20:22 | |
| *** jasondotstar has quit IRC | 20:22 | |
| *** jogo has quit IRC | 20:25 | |
| melwitt | mriedem: where is the thing where you see mox not verifying order of calls? I also thought it does | 20:25 |
| *** jogo has joined #openstack-nova | 20:26 | |
| mriedem | melwitt: in http://git.openstack.org/cgit/openstack/nova/tree/nova/tests/unit/network/test_manager.py#n922 | 20:26 |
| mriedem | melwitt: i have a change where virtual_interface_get_by_instance_and_network is called before fixed_ip_associate_pool | 20:26 |
| mriedem | but the test doesn't fail | 20:26 |
| *** toure has quit IRC | 20:29 | |
| *** Nikolay_St has quit IRC | 20:30 | |
| melwitt | mriedem: yeah, I thought that should fail. only thing I can think so far is if somehow the mox from mox3 is different, since that's where mox is coming from in oslo.test | 20:30 |
| mriedem | melwitt: yeah, that's what i was wondering | 20:31 |
| mriedem | https://github.com/openstack/oslotest/commit/9e0c8ad2c251274128499a7fcfb591c488d27d2b | 20:31 |
| *** baoli has quit IRC | 20:31 | |
| *** dontalton has quit IRC | 20:32 | |
| *** dontalton has joined #openstack-nova | 20:33 | |
| *** greghaynes has joined #openstack-nova | 20:33 | |
| *** greghaynes has quit IRC | 20:33 | |
| *** claudiub has quit IRC | 20:34 | |
| *** nic has quit IRC | 20:34 | |
| *** haypo has joined #openstack-nova | 20:38 | |
| *** vladikr has quit IRC | 20:39 | |
| *** rwsu has joined #openstack-nova | 20:39 | |
| *** ajayaa has quit IRC | 20:41 | |
| melwitt | mriedem: maybe you already saw this but it intends to verify ordering from the look of the tests, I don't see a test that verifies it raises UnexpectedMethodCallError if calls are made out of order though https://github.com/openstack/mox3/blob/master/mox3/tests/test_mox.py other than that you can sanity test substituting to use plain mox and see what it does. the docs say it's strict about ordering and that's been my experience with it | 20:43 |
| *** claudiub has joined #openstack-nova | 20:44 | |
| mriedem | melwitt: i see a few self.assertRaises(mox.UnexpectedMethodCallError, call) in there | 20:45 |
| mriedem | melwitt: i could maybe poke that bear later, for now i just wrote new tests | 20:45 |
| mriedem | since this has to be backported anyway | 20:45 |
| *** Nikolay_St has joined #openstack-nova | 20:46 | |
| melwitt | mriedem: I thought those were boolean like called when not expected to be called, but I could have missed something | 20:46 |
| melwitt | mriedem: okay, cool | 20:47 |
| *** promulo has quit IRC | 20:48 | |
| *** annashen has joined #openstack-nova | 20:51 | |
| *** htruta has quit IRC | 20:52 | |
| bauzas | melwitt: hey, while you're here, I have a question for you :) | 20:54 |
| bauzas | melwitt: https://bugs.launchpad.net/nova/+bug/1468087 | 20:54 |
| openstack | Launchpad bug 1468087 in OpenStack Compute (nova) "py27 job failing on test_save_objectfield_missing_instance_row" [High,Confirmed] | 20:54 |
| bauzas | melwitt: for some reason the unittest you added is failing since 24h | 20:54 |
| bauzas | melwitt: tbc, it's not failing all of times, but rather racing | 20:55 |
| bauzas | melwitt: and on the remote side only | 20:55 |
| dims | jogo: thanks | 20:55 |
| *** ivasev has quit IRC | 20:56 | |
| *** changbl_ has quit IRC | 20:56 | |
| *** alexpilotti has quit IRC | 20:57 | |
| *** scheuran has quit IRC | 20:58 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Reduce window for allocate_fixed_ip / release_fixed_ip race in nova-net https://review.openstack.org/194815 | 20:58 |
| mriedem | dansmith: melwitt: sdague: that's part one ^ | 20:58 |
| bauzas | melwitt: like if the mocks were not working | 20:58 |
| bauzas | melwitt: unfortunately, we don't know which field method is ok | 20:59 |
| *** ociuhandu has joined #openstack-nova | 20:59 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova-specs: Add rootwrap-daemon-mode blueprint https://review.openstack.org/105404 | 21:00 |
| *** Nikolay_St has quit IRC | 21:02 | |
| *** Raildo_ has joined #openstack-nova | 21:03 | |
| bauzas | melwitt: no worries, I'm going off, I'll be looking at the problem by tomorrow | 21:03 |
| haypo | jogo, hi. for https://review.openstack.org/#/c/194128/ "Make gate-nova-python34 voting" i have no strong opinion on adding a py34 gate, not only py34 check. the first version of my spec added both jobs | 21:03 |
| *** annegentle has quit IRC | 21:03 | |
| *** chen12 has quit IRC | 21:04 | |
| haypo | i'm now unable to find logs of the discussion on the py34 gate. maybe it was on irc, sorry i don't remember | 21:04 |
| *** chen12 has joined #openstack-nova | 21:04 | |
| *** lyarwood has quit IRC | 21:04 | |
| *** lyarwood has joined #openstack-nova | 21:05 | |
| dansmith | alaski: oops, sorry I missed your comments on the migration status patch | 21:07 |
| dansmith | alaski: no vote meant it didn't get seen | 21:07 |
| dansmith | alaski: I'll circle bac | 21:07 |
| dansmith | k | 21:07 |
| *** Nikolay_1t has joined #openstack-nova | 21:07 | |
| *** alex_klimov has joined #openstack-nova | 21:07 | |
| dims | jogo: https://review.openstack.org/105404 is ready! thanks | 21:07 |
| alaski | dansmith: no worries. It wasn't quite enough for a -1 | 21:07 |
| *** lpetrut has quit IRC | 21:08 | |
| *** eharney has quit IRC | 21:09 | |
| *** radek__ has quit IRC | 21:10 | |
| melwitt | bauzas: looking into it, I forgot I added that test | 21:10 |
| *** aysyanne has quit IRC | 21:10 | |
| *** Marga_ has quit IRC | 21:10 | |
| jogo | haypo: without the gate job we can and will wedge things | 21:11 |
| jogo | an easy example would be a patch that expands the set of tests to run under py34 | 21:11 |
| haypo | jogo, the rationale was "it's unlikely" :-) | 21:11 |
| *** adrian_otto has joined #openstack-nova | 21:11 | |
| ctrath | mriedem: so, how does moving the vif call to the beginning of the try block reduce the race window? | 21:11 |
| jogo | haypo: with the number of patches we approve daily unlikely will happen | 21:11 |
| haypo | jogo, ok. i don't think that it will add so much workload to the openstack infra | 21:12 |
| mriedem | ctrath: b/c today the order is associate / allocate vif | 21:13 |
| mriedem | ctrath: the release_fixed_ip method is checking association and allocated flags | 21:13 |
| ctrath | mriedem: I'm assuming that get_by_instance_and_network just adds padded time before getting the address | 21:13 |
| mriedem | ctrath: so we want to reduce any amount of time it takes to set the allocated flag to True | 21:13 |
| mriedem | ctrath: there is also _add_virtual_interface | 21:13 |
| jogo | haypo: if we hit any resource issues we can revisit this yeah | 21:13 |
| openstackgerrit | dane-fichter proposed openstack/nova-specs: Nova Support of Glance Image Signing https://review.openstack.org/188874 | 21:14 |
| haypo | jogo, it sounds like a good plan. in my short experience of openstack, it's really annoying when a gate is broken (for any reason) | 21:14 |
| ctrath | mriedem: ok.. I'm also assuming that the real fix in release_fixed_ip is going to take time and investigation, thus the reason for this "hack"? | 21:15 |
| openstackgerrit | dane-fichter proposed openstack/nova-specs: Nova Support of Glance Image Signing https://review.openstack.org/188874 | 21:15 |
| mriedem | ctrath: we know what the real fix is and i'm working on that | 21:15 |
| mriedem | ctrath: but it requires rpc api version changes which we generally don't want to backport to stable branches | 21:15 |
| ctrath | gotcha.. thanks | 21:15 |
| mriedem | so the first patch is a tactical backportable patch for stable | 21:15 |
| *** rlrossit has left #openstack-nova | 21:15 | |
| mriedem | then we do the rpc api version change | 21:15 |
| *** annegentle has joined #openstack-nova | 21:16 | |
| *** fawadkhaliq has joined #openstack-nova | 21:17 | |
| ctrath | mriedem: is there a way to test to see if the address is not yet released? | 21:17 |
| jogo | haypo: yup, it makes developers stabby | 21:17 |
| mriedem | ctrath: so the dhcpagent callback is what's calling release_fixed_ip | 21:18 |
| mriedem | so that's on a different thread | 21:18 |
| *** karimb has joined #openstack-nova | 21:18 | |
| mriedem | ctrath: so the fix is going to be to see if the released mac is the same as the vif that is associated with the instance that is tied to the fixed IP being released, | 21:18 |
| mriedem | if not, it's the wrong vif and the wrong instance and we should ignore release_fixed_ip | 21:18 |
| ctrath | ok.. That's fine.. I'm wondering if we could try to get the address in a for loop until we successfully obtain it, or fail at a given timeout | 21:19 |
| *** matrohon has joined #openstack-nova | 21:19 | |
| *** tonytan4ever has joined #openstack-nova | 21:20 | |
| mriedem | ctrath: well, there is another way to fix this, which is making the association between the fixed_ip and instance atomic with setting the fixed_ip.allocated flag to True | 21:20 |
| mriedem | in the db api | 21:20 |
| mriedem | b/c then release_fixed_ip would see that the fixed IP is allocated and won't disassociate it | 21:20 |
| *** fawadkhaliq has quit IRC | 21:21 | |
| mriedem | we could potentially do both changes, but both require rpc api version changes so neither are backportable really | 21:21 |
| *** oro has joined #openstack-nova | 21:22 | |
| ctrath | yeah, Ok. I was trying to see if there was a better way of adverting this other than lengthining the time window | 21:22 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova: Switch to using os-brick https://review.openstack.org/175569 | 21:22 |
| mriedem | ctrath: we're fixing release_fixed_ip first b/c it's dangerous on it's own given it can potentially disassociate fixed IPs from the wrong instance - we need the mac comparisong regardless | 21:22 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova: Add os_brick-based VolumeDriver for HGST connector https://review.openstack.org/194412 | 21:24 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova: Add os-brick based LibvirtVolumeDriver for ScaleIO https://review.openstack.org/194454 | 21:24 |
| *** thorst has quit IRC | 21:26 | |
| *** Nikolay_1t has quit IRC | 21:27 | |
| melwitt | dansmith: I noticed on your patch that fixes the test_save_objectfield_missing_instance_row, the serialize_entity gets stuck in a loop doing the remote test, do you understand it? http://logs.openstack.org/05/192805/3/gate/gate-nova-python27/da81c4f/testr_results.html.gz | 21:28 |
| *** matrohon has quit IRC | 21:28 | |
| dansmith | melwitt: hmm, I saw that once locally and then never again after repeated attempts | 21:28 |
| dansmith | so, no... I'll have to think on it a bit | 21:28 |
| *** krtaylor has quit IRC | 21:29 | |
| melwitt | dansmith: okay. I'll recheck it then, I think your patch will fix the bug that bauzas is looking into with that unit test failing sometimes (because of attr not set) | 21:29 |
| dansmith | well, we don't want to merge this if it's doing this | 21:30 |
| dansmith | I just didn't see it again, but it's clearly still a problem | 21:30 |
| *** adrian_otto has quit IRC | 21:30 | |
| melwitt | okay | 21:31 |
| *** romainh has quit IRC | 21:32 | |
| *** Nikolay_St has joined #openstack-nova | 21:33 | |
| *** annegentle has quit IRC | 21:34 | |
| *** annegentle has joined #openstack-nova | 21:34 | |
| *** iamjarvo has quit IRC | 21:34 | |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova: Switch to using os-brick https://review.openstack.org/175569 | 21:40 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova: Add os_brick-based VolumeDriver for HGST connector https://review.openstack.org/194412 | 21:40 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova: Add os-brick based LibvirtVolumeDriver for ScaleIO https://review.openstack.org/194454 | 21:40 |
| openstackgerrit | Tracy Jones proposed openstack/nova: Cleanup of compute api reboot method https://review.openstack.org/127427 | 21:43 |
| *** kentb1 has joined #openstack-nova | 21:44 | |
| *** electrocucaracha has quit IRC | 21:46 | |
| melwitt | dansmith: looking at it, I'm guessing it's because the test is manipulating _changed_fields without actually setting the field to anything, and might be messing up the serialize logic. ideally, the test should just set the field to something instead of adding to changed fields | 21:46 |
| *** kentb1 has quit IRC | 21:48 | |
| dansmith | hmm, I dunno it's trying to load ec2_ids | 21:48 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/194325 | 21:48 |
| *** electrocucaracha has joined #openstack-nova | 21:49 | |
| dansmith | oh | 21:50 |
| *** annashen has quit IRC | 21:51 | |
| dansmith | so actually I should just make that the list of things that touch instance_extra | 21:51 |
| *** jecarey has quit IRC | 21:51 | |
| dansmith | just because they are a _save_thing() doesn't mean they should be in this test | 21:51 |
| dansmith | and ec2_ids is one of those | 21:51 |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Adds Hyper-V Shielded VMs spec https://review.openstack.org/192892 | 21:53 |
| *** thedodd has quit IRC | 21:53 | |
| melwitt | dansmith: hm, okay | 21:53 |
| dansmith | or just move this test to the non-remote class | 21:54 |
| dansmith | because it's mocking out things that make sense for us to do, but that are just confusing the serialization stuff | 21:54 |
| *** ianbrown has joined #openstack-nova | 21:54 | |
| *** Nikolay_St has quit IRC | 21:54 | |
| *** haypo has quit IRC | 21:55 | |
| melwitt | dansmith: it's just because there isn't any ec2 ids set for the fake instance then? yeah, I wasn't think of remote when I wrote that | 21:55 |
| dansmith | well, | 21:55 |
| dansmith | because it's not set, but we're telling the thing that it _is_set | 21:56 |
| dansmith | so the serialization stuff wants to serialize it since it's set | 21:56 |
| dansmith | but it's not, so the getattr triggers a load | 21:56 |
| melwitt | ah, yeah | 21:56 |
| dansmith | so I'll just move it | 21:56 |
| melwitt | okay, cool. makes sense | 21:57 |
| dansmith | incoming | 21:58 |
| openstackgerrit | Dan Smith proposed openstack/nova: Remove compat_instance() https://review.openstack.org/192708 | 21:58 |
| openstackgerrit | Dan Smith proposed openstack/nova: Fix testing object fields with missing instance rows https://review.openstack.org/192805 | 21:58 |
| *** marun has quit IRC | 21:59 | |
| *** oro has quit IRC | 21:59 | |
| *** chen12 has quit IRC | 22:00 | |
| *** Nikolay_St has joined #openstack-nova | 22:00 | |
| openstackgerrit | Joe Gordon proposed openstack/nova: Replace ascii art architecture diagram with svg image https://review.openstack.org/194840 | 22:01 |
| *** agireud has quit IRC | 22:01 | |
| jogo | mriedem: ^ | 22:02 |
| *** burt has quit IRC | 22:03 | |
| mtreinish | jogo: heh, https://review.openstack.org/#/c/194840/1/doc/source/images/architecture.svg is totally hand editable in an editor :) | 22:03 |
| jogo | ol | 22:04 |
| jogo | lol | 22:04 |
| harlowja | good think that dia stuff is opensource :-P | 22:04 |
| *** otter768 has joined #openstack-nova | 22:04 | |
| harlowja | *thing | 22:04 |
| jogo | too bad the dia file isn't | 22:04 |
| harlowja | * https://review.openstack.org/cat/194840%2C1%2Cdoc/source/images/architecture.svg%5E0 (for those that want to view it) | 22:04 |
| *** claudiub has quit IRC | 22:05 | |
| harlowja | should render in all (less than 5 year old) browsers | 22:05 |
| *** edmondsw has quit IRC | 22:05 | |
| jogo | harlowja: neat! how did you do that | 22:05 |
| *** electrocucaracha has quit IRC | 22:05 | |
| jogo | err find it | 22:05 |
| harlowja | jogo http://i.imgur.com/Yph3auH.png | 22:06 |
| harlowja | click on that download link, and thats ^ | 22:06 |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Check mac for instance before disassociate in release_fixed_ip https://review.openstack.org/194843 | 22:06 |
| harlowja | the little arrow thingy | 22:06 |
| *** electrocucaracha has joined #openstack-nova | 22:06 | |
| jogo | ahh slick | 22:06 |
| *** karimb has quit IRC | 22:06 | |
| mtreinish | harlowja: heh, I was going to make a joke about using lynx but it did the right thing and asked would you like to download the image file | 22:07 |
| jogo | haha | 22:07 |
| harlowja | non-terminal based, (less than 5 year old) browsers | 22:07 |
| harlowja | * http://caniuse.com/#search=svg | 22:07 |
| harlowja | all those, ha | 22:07 |
| jogo | harlowja: so just not IE8 | 22:08 |
| harlowja | seems so | 22:08 |
| harlowja | oh well | 22:08 |
| *** annegent_ has joined #openstack-nova | 22:09 | |
| *** otter768 has quit IRC | 22:09 | |
| harlowja | and some old andriod browsers (that apparently have partial support) | 22:09 |
| mtreinish | harlowja: that site doesn't even mention lynx... :( | 22:09 |
| harlowja | lol | 22:09 |
| harlowja | https://lists.nongnu.org/archive/html/lynx-dev/2010-08/msg00007.html ( SVG support in Lynx) | 22:10 |
| harlowja | lol | 22:10 |
| harlowja | so who knows, something may work there :-P | 22:10 |
| *** Raildo_ has quit IRC | 22:11 | |
| *** xgsa has quit IRC | 22:11 | |
| *** annegentle has quit IRC | 22:13 | |
| *** annashen has joined #openstack-nova | 22:13 | |
| *** annashen has quit IRC | 22:13 | |
| *** annashen has joined #openstack-nova | 22:13 | |
| *** chen12 has joined #openstack-nova | 22:15 | |
| *** xgsa has joined #openstack-nova | 22:17 | |
| * dansmith channels mriedem | 22:17 | |
| dansmith | gd pep8 | 22:17 |
| *** Sukhdev has quit IRC | 22:19 | |
| *** dontalton2 has joined #openstack-nova | 22:19 | |
| *** mriedem has quit IRC | 22:20 | |
| *** husanu has quit IRC | 22:20 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Enable python34 tests for nova/tests/unit/objects/test*.py https://review.openstack.org/187785 | 22:20 |
| *** Nikolay_St has quit IRC | 22:20 | |
| *** Nikolay_St has joined #openstack-nova | 22:21 | |
| *** jgrimm is now known as zz_jgrimm | 22:21 | |
| *** rfolco has quit IRC | 22:21 | |
| *** annegent_ has quit IRC | 22:22 | |
| *** husanu has joined #openstack-nova | 22:22 | |
| *** annegentle has joined #openstack-nova | 22:22 | |
| *** gyee has quit IRC | 22:23 | |
| *** yamahata has quit IRC | 22:23 | |
| dims | dansmith: seen the test_save_objectfield_missing_instance_row intermittent error yet? | 22:25 |
| dansmith | dims: in my patch or in general? | 22:25 |
| dims | dansmith: general i think now - https://bugs.launchpad.net/nova/+bug/1468087 | 22:26 |
| openstack | Launchpad bug 1468087 in OpenStack Compute (nova) "py27 job failing on test_save_objectfield_missing_instance_row" [High,Confirmed] | 22:26 |
| dansmith | dims: I see your bug and raise you a patch: https://review.openstack.org/#/c/192805/ | 22:26 |
| *** xyang1 has quit IRC | 22:26 | |
| dims | dansmith: yay! | 22:27 |
| *** gyee has joined #openstack-nova | 22:27 | |
| tjones1 | melwitt or dansmith: i've modifed part of manage.py to convert to using objects. Im looking for the unit tests to run (since i don't want to spend 2 hours running them all). For example, I grepped for VpnCommands to see if i could find the tests and i don't see any. Are there unit tests for nova/cmd/manage.py? if you don't know off hand i can keep looking (or run all tests) | 22:28 |
| dansmith | tjones1: nova/tests/unit/test_manage.py? | 22:28 |
| *** alex_klimov has quit IRC | 22:28 | |
| tjones1 | dansmith: so simple | 22:28 |
| tjones1 | thanks | 22:28 |
| tjones1 | let me look for that object | 22:29 |
| openstackgerrit | Dan Smith proposed openstack/nova: Remove compat_instance() https://review.openstack.org/192708 | 22:29 |
| openstackgerrit | Dan Smith proposed openstack/nova: Fix testing object fields with missing instance rows https://review.openstack.org/192805 | 22:29 |
| tjones1 | dansmith: ah no wonder. no tests for VpnCommands | 22:29 |
| tjones1 | thanks for the pointer | 22:30 |
| dansmith | np | 22:30 |
| dims | dansmith: need 2 obj_reset_changes? one before and one after save? | 22:31 |
| *** ijuwang is now known as ijuwang_away | 22:32 | |
| *** dboik has quit IRC | 22:32 | |
| dims | dansmith: ah gotcha we are using the same instance and testing all the fields. nm | 22:32 |
| *** kylek3h has quit IRC | 22:32 | |
| *** gholler has quit IRC | 22:34 | |
| *** annashen has quit IRC | 22:35 | |
| melwitt | dims: I'm still trying to understand what is going on in that failure, what "mismatch method returned none" means | 22:37 |
| melwitt | dims: I think dansmith's patch will fix it, but I'm not clear on exactly what's wrong yet | 22:38 |
| dims | melwitt: save was supposed to throw an exception and it doesn't | 22:38 |
| dansmith | no | 22:38 |
| dims | it returns None instead | 22:38 |
| melwitt | oh /facepalm | 22:39 |
| dansmith | well, the bug is actually a different thing | 22:39 |
| dansmith | but it's probably related | 22:39 |
| dansmith | this moves it out of the Remote class anyway, which I'm sure will fix it | 22:39 |
| *** zzzeek has quit IRC | 22:40 | |
| dansmith | the test is mucking with stuff that is going to break the remoting regardless | 22:40 |
| dims | dansmith: right, was explaining what the testtools.matchers._impl.MismatchError is tell us | 22:40 |
| dims | telling us | 22:40 |
| dansmith | dims: oh right, sorry | 22:40 |
| * dansmith retracts his no | 22:40 | |
| dims | :) | 22:40 |
| *** ctrath has quit IRC | 22:42 | |
| *** mdrabe has quit IRC | 22:44 | |
| openstackgerrit | Tracy Jones proposed openstack/nova: WIP - convert direct db calls to objects https://review.openstack.org/194851 | 22:45 |
| *** rwsu has quit IRC | 22:47 | |
| *** Nikolay_St has quit IRC | 22:48 | |
| *** Nikolay_St has joined #openstack-nova | 22:48 | |
| melwitt | dansmith: I wonder if it's because the test mock.patch.object(instance) instead of mock.patch('nova.objects.Instance.save...') so the exception doesn't get raised after it's made the remote trip | 22:52 |
| *** IanGovett has quit IRC | 22:54 | |
| *** baoli has joined #openstack-nova | 22:58 | |
| *** annashen has joined #openstack-nova | 22:58 | |
| *** baoli has quit IRC | 22:59 | |
| *** baoli has joined #openstack-nova | 23:00 | |
| *** r-daneel has quit IRC | 23:00 | |
| dansmith | melwitt: could be yeah, but it should be repeatable that way I'd think | 23:03 |
| *** annegentle has quit IRC | 23:03 | |
| *** annegentle has joined #openstack-nova | 23:05 | |
| *** artom has quit IRC | 23:05 | |
| *** mriedem has joined #openstack-nova | 23:05 | |
| mriedem | dansmith: gd is right | 23:05 |
| dansmith | heh | 23:06 |
| *** baoli has quit IRC | 23:06 | |
| *** artom has joined #openstack-nova | 23:06 | |
| *** sdake has joined #openstack-nova | 23:07 | |
| dansmith | mriedem: +2 this: https://review.openstack.org/#/c/192805/ | 23:07 |
| *** tonytan4ever has quit IRC | 23:08 | |
| dansmith | it happens to fix https://bugs.launchpad.net/nova/+bug/1468087 | 23:08 |
| openstack | Launchpad bug 1468087 in OpenStack Compute (nova) "py27 job failing on test_save_objectfield_missing_instance_row" [High,Confirmed] | 23:08 |
| *** annegentle has quit IRC | 23:10 | |
| mriedem | looking | 23:11 |
| *** dboik has joined #openstack-nova | 23:12 | |
| *** ijw_ has joined #openstack-nova | 23:13 | |
| *** dboik has quit IRC | 23:14 | |
| mriedem | +Walrus | 23:15 |
| dansmith | +Woohoo | 23:15 |
| *** Nikolay_St has quit IRC | 23:15 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Fix testing object fields with missing instance rows https://review.openstack.org/192805 | 23:15 |
| mriedem | dansmith: and then i went and f'ed up your topic branch | 23:16 |
| mriedem | added the bug tag to the commit message | 23:16 |
| *** Nikolay_St has joined #openstack-nova | 23:16 | |
| dansmith | heh, okay | 23:16 |
| *** dontalton2 has quit IRC | 23:16 | |
| *** dontalton has quit IRC | 23:16 | |
| dansmith | it didn't set out to fix the bug, it's just a happy coincidence | 23:16 |
| dansmith | but as you wish | 23:16 |
| *** lucas-dinner has quit IRC | 23:17 | |
| kfox1111 | johnthetubaguy: how are we doing with the nova instance user spec for the spec freeze? the barbican and keystone ptl's have both +1'ed it. | 23:20 |
| *** erlon has quit IRC | 23:20 | |
| *** angdraug has quit IRC | 23:21 | |
| mriedem | it's going to be a terrible couple of weeks for johnthetubaguy | 23:21 |
| kfox1111 | heh. yeah. | 23:22 |
| *** mgarza_ has quit IRC | 23:22 | |
| kfox1111 | and for us folks trying so hard not to have to wait another 6 months. :/ | 23:23 |
| *** sdake_ has joined #openstack-nova | 23:23 | |
| *** jasondotstar has joined #openstack-nova | 23:23 | |
| mriedem | i know, it's all terrible | 23:24 |
| openstackgerrit | Michael Still proposed openstack/nova: libvirt: Kill rsync/scp processes before deleting instance https://review.openstack.org/192986 | 23:24 |
| *** otter768 has joined #openstack-nova | 23:25 | |
| *** VW_ has quit IRC | 23:26 | |
| *** sdake has quit IRC | 23:26 | |
| *** ijw_ has quit IRC | 23:28 | |
| *** ijw_ has joined #openstack-nova | 23:28 | |
| *** ZZelle has quit IRC | 23:35 | |
| *** hemna is now known as hemnafk | 23:37 | |
| *** mriedem has quit IRC | 23:38 | |
| *** mriedem has joined #openstack-nova | 23:39 | |
| *** sdake_ has quit IRC | 23:40 | |
| *** Nikolay_St has quit IRC | 23:42 | |
| *** annashen has quit IRC | 23:42 | |
| *** Nikolay_St has joined #openstack-nova | 23:43 | |
| *** tjones1 has quit IRC | 23:47 | |
| *** isq has quit IRC | 23:54 | |
| *** dims has quit IRC | 23:56 | |
| *** ijw_ has quit IRC | 23:58 | |
| *** moshele_ has joined #openstack-nova | 23:58 | |
| *** moshele_ has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
