| *** dimsum__ has joined #openstack-neutron | 00:06 | |
| *** sdake_ has joined #openstack-neutron | 00:09 | |
| *** sdake has quit IRC | 00:12 | |
| *** merooney has joined #openstack-neutron | 00:14 | |
| *** dimsum__ has quit IRC | 00:14 | |
| *** merooney has quit IRC | 00:18 | |
| *** nlahouti has joined #openstack-neutron | 00:18 | |
| *** yamamoto has joined #openstack-neutron | 00:18 | |
| *** amotoki has joined #openstack-neutron | 00:19 | |
| *** nlahouti has quit IRC | 00:20 | |
| *** nlahouti has joined #openstack-neutron | 00:21 | |
| *** merooney has joined #openstack-neutron | 00:23 | |
| *** jogo has quit IRC | 00:25 | |
| *** tkatarki has joined #openstack-neutron | 00:35 | |
| *** tkatarki has quit IRC | 00:37 | |
| *** tkatarki has joined #openstack-neutron | 00:37 | |
| *** dimsum__ has joined #openstack-neutron | 00:39 | |
| *** ivar-laz_ has joined #openstack-neutron | 00:40 | |
| *** ivar-la__ has joined #openstack-neutron | 00:42 | |
| *** ivar-lazzaro has quit IRC | 00:42 | |
| *** ivar-laz_ has quit IRC | 00:42 | |
| *** shwetaap has joined #openstack-neutron | 00:50 | |
| *** gildub has joined #openstack-neutron | 00:51 | |
| *** yamahata has quit IRC | 00:55 | |
| *** roeyc has quit IRC | 00:56 | |
| *** julim has joined #openstack-neutron | 01:05 | |
| openstackgerrit | Ian Wienand proposed openstack/neutron: Ensure netfilter is enabled for bridges https://review.openstack.org/180867 | 01:05 |
|---|---|---|
| *** merooney has quit IRC | 01:08 | |
| *** merooney has joined #openstack-neutron | 01:08 | |
| *** yuanying has joined #openstack-neutron | 01:10 | |
| *** yalie has joined #openstack-neutron | 01:16 | |
| *** mturpin has quit IRC | 01:18 | |
| *** tkatarki has quit IRC | 01:20 | |
| *** yalie has left #openstack-neutron | 01:24 | |
| *** liusheng has quit IRC | 01:27 | |
| *** liusheng has joined #openstack-neutron | 01:31 | |
| *** erkules has joined #openstack-neutron | 01:35 | |
| *** markvoelker has quit IRC | 01:35 | |
| *** markvoelker has joined #openstack-neutron | 01:36 | |
| *** banix has quit IRC | 01:37 | |
| *** markvoelker has quit IRC | 01:37 | |
| *** erkules_ has quit IRC | 01:38 | |
| *** baoli has joined #openstack-neutron | 01:39 | |
| *** baoli has quit IRC | 01:44 | |
| *** aranjan has joined #openstack-neutron | 01:47 | |
| *** julim has quit IRC | 01:48 | |
| *** shwetaap has quit IRC | 01:53 | |
| *** achanda has joined #openstack-neutron | 01:55 | |
| *** banix has joined #openstack-neutron | 01:56 | |
| *** shwetaap has joined #openstack-neutron | 02:06 | |
| *** dimsum__ has quit IRC | 02:12 | |
| *** jogo has joined #openstack-neutron | 02:22 | |
| *** yasu_ has joined #openstack-neutron | 02:22 | |
| *** park_heijlong has joined #openstack-neutron | 02:22 | |
| *** shwetaap has quit IRC | 02:23 | |
| *** shwetaap has joined #openstack-neutron | 02:29 | |
| *** zhhuabj_ has joined #openstack-neutron | 02:30 | |
| *** zhhuabj has quit IRC | 02:32 | |
| *** Carlito__ has joined #openstack-neutron | 02:37 | |
| *** yamamoto has quit IRC | 02:40 | |
| *** Carlito__ has quit IRC | 02:41 | |
| *** britthou_ has joined #openstack-neutron | 02:41 | |
| *** britthouser has quit IRC | 02:41 | |
| *** yamamoto has joined #openstack-neutron | 02:41 | |
| *** Carlito__ has joined #openstack-neutron | 02:42 | |
| *** Carlito__ has quit IRC | 02:42 | |
| *** amotoki has quit IRC | 02:43 | |
| *** yamamoto_ has joined #openstack-neutron | 02:49 | |
| *** yamamoto has quit IRC | 02:52 | |
| *** merooney has quit IRC | 02:54 | |
| *** amotoki has joined #openstack-neutron | 02:56 | |
| *** boris-42 has quit IRC | 02:58 | |
| *** dave-mccowan has quit IRC | 03:00 | |
| *** achanda has quit IRC | 03:01 | |
| *** fedexo has quit IRC | 03:04 | |
| *** achanda_ has joined #openstack-neutron | 03:04 | |
| *** fedexo has joined #openstack-neutron | 03:05 | |
| *** sdake has joined #openstack-neutron | 03:05 | |
| *** sdake_ has quit IRC | 03:06 | |
| *** sdake has quit IRC | 03:08 | |
| *** sdake has joined #openstack-neutron | 03:09 | |
| openstackgerrit | Ian Wienand proposed openstack/neutron: Ensure netfilter is enabled for bridges https://review.openstack.org/180867 | 03:09 |
| *** achanda_ has quit IRC | 03:14 | |
| *** tobe has joined #openstack-neutron | 03:20 | |
| *** alex_xu has quit IRC | 03:20 | |
| *** baoli has joined #openstack-neutron | 03:21 | |
| *** alex_xu has joined #openstack-neutron | 03:21 | |
| *** yamamoto_ has quit IRC | 03:22 | |
| *** shashankhegde has joined #openstack-neutron | 03:24 | |
| *** baoli has quit IRC | 03:26 | |
| *** subscope has joined #openstack-neutron | 03:38 | |
| *** shwetaap has quit IRC | 03:42 | |
| *** yamamoto has joined #openstack-neutron | 03:44 | |
| *** banix has quit IRC | 03:51 | |
| *** merooney has joined #openstack-neutron | 03:54 | |
| *** sdake_ has joined #openstack-neutron | 03:56 | |
| *** carl_baldwin has joined #openstack-neutron | 03:57 | |
| *** rkukura has quit IRC | 03:59 | |
| *** rkukura has joined #openstack-neutron | 03:59 | |
| *** vikram has joined #openstack-neutron | 04:00 | |
| *** merooney has quit IRC | 04:00 | |
| *** sdake has quit IRC | 04:00 | |
| *** vikram has quit IRC | 04:03 | |
| *** yatin has joined #openstack-neutron | 04:04 | |
| openstackgerrit | shihanzhang proposed openstack/neutron: setup port filters when sg rules change https://review.openstack.org/181272 | 04:14 |
| *** coolsvap|afk is now known as coolsvap | 04:23 | |
| *** yamamoto has quit IRC | 04:28 | |
| *** yamamoto has joined #openstack-neutron | 04:30 | |
| *** sdake has joined #openstack-neutron | 04:34 | |
| *** trinaths has joined #openstack-neutron | 04:37 | |
| *** ivar-la__ has quit IRC | 04:37 | |
| *** markvoelker has joined #openstack-neutron | 04:37 | |
| *** ivar-lazzaro has joined #openstack-neutron | 04:38 | |
| *** sdake_ has quit IRC | 04:38 | |
| *** trinaths has left #openstack-neutron | 04:44 | |
| *** ivar-lazzaro has quit IRC | 04:45 | |
| *** ivar-laz_ has joined #openstack-neutron | 04:45 | |
| *** irenab has quit IRC | 04:46 | |
| *** carl_baldwin has quit IRC | 04:48 | |
| *** afazekas has joined #openstack-neutron | 04:55 | |
| *** yamamoto has quit IRC | 04:57 | |
| openstackgerrit | Ian Wienand proposed openstack/neutron: Ensure netfilter is enabled for bridges with OVS-Hybrid Driver https://review.openstack.org/180867 | 04:59 |
| *** tobe has quit IRC | 05:03 | |
| *** red_trela has joined #openstack-neutron | 05:07 | |
| *** chlong has quit IRC | 05:09 | |
| *** neeti has joined #openstack-neutron | 05:09 | |
| *** sdake_ has joined #openstack-neutron | 05:13 | |
| *** emagana has joined #openstack-neutron | 05:15 | |
| *** sdake has quit IRC | 05:16 | |
| *** yamamoto has joined #openstack-neutron | 05:17 | |
| *** tflynn has joined #openstack-neutron | 05:22 | |
| *** trinaths has joined #openstack-neutron | 05:22 | |
| *** tflynn_ has joined #openstack-neutron | 05:23 | |
| *** daneyon has joined #openstack-neutron | 05:25 | |
| *** tflynn has quit IRC | 05:27 | |
| *** heyongli has quit IRC | 05:29 | |
| *** ekuris has joined #openstack-neutron | 05:30 | |
| *** irenab has joined #openstack-neutron | 05:30 | |
| *** yamamoto has quit IRC | 05:30 | |
| *** heyongli has joined #openstack-neutron | 05:31 | |
| *** gildub has quit IRC | 05:33 | |
| *** yamamoto has joined #openstack-neutron | 05:33 | |
| *** achanda has joined #openstack-neutron | 05:37 | |
| *** yamamoto has quit IRC | 05:37 | |
| openstackgerrit | Ian Wienand proposed openstack/neutron: Ensure netfilter is enabled for bridges with OVS-Hybrid Driver https://review.openstack.org/180867 | 05:38 |
| *** jogo has quit IRC | 05:39 | |
| *** jogo has joined #openstack-neutron | 05:40 | |
| *** vthapar has joined #openstack-neutron | 05:40 | |
| *** anshul has joined #openstack-neutron | 05:41 | |
| *** yfauser has joined #openstack-neutron | 05:42 | |
| *** yfauser has quit IRC | 05:43 | |
| *** yamamoto has joined #openstack-neutron | 05:47 | |
| *** shashankhegde has quit IRC | 05:49 | |
| *** itzikb has joined #openstack-neutron | 05:54 | |
| *** sdake has joined #openstack-neutron | 05:57 | |
| *** sudipto has joined #openstack-neutron | 05:58 | |
| *** shashankhegde has joined #openstack-neutron | 05:59 | |
| *** Longgeek_ has joined #openstack-neutron | 05:59 | |
| *** sdake_ has quit IRC | 06:01 | |
| *** Longgeek has quit IRC | 06:03 | |
| *** nlahouti has quit IRC | 06:03 | |
| *** luqas has joined #openstack-neutron | 06:04 | |
| *** yamamoto has quit IRC | 06:06 | |
| *** yamamoto has joined #openstack-neutron | 06:08 | |
| openstackgerrit | Qin Zhao proposed openstack/python-neutronclient: Add InvalidIpForSubnetClient exception https://review.openstack.org/162062 | 06:08 |
| *** luqas has quit IRC | 06:09 | |
| *** enikanorov2 has quit IRC | 06:10 | |
| *** yalie has joined #openstack-neutron | 06:11 | |
| *** simon-AS559 has quit IRC | 06:12 | |
| *** anilvenkata has joined #openstack-neutron | 06:14 | |
| *** sdake_ has joined #openstack-neutron | 06:15 | |
| *** red_trela has quit IRC | 06:16 | |
| *** sdake__ has joined #openstack-neutron | 06:18 | |
| *** sdake has quit IRC | 06:18 | |
| *** sdake_ has quit IRC | 06:21 | |
| *** red_trela has joined #openstack-neutron | 06:23 | |
| *** yfried_ is now known as yfried|afk | 06:24 | |
| *** yfried|afk is now known as yfried_ | 06:24 | |
| *** simon-AS559 has joined #openstack-neutron | 06:26 | |
| *** scheuran has joined #openstack-neutron | 06:26 | |
| *** numan has joined #openstack-neutron | 06:28 | |
| *** red_trela has quit IRC | 06:32 | |
| *** nmagnezi has joined #openstack-neutron | 06:33 | |
| *** yog__ has joined #openstack-neutron | 06:35 | |
| *** Murali has joined #openstack-neutron | 06:38 | |
| *** red_trela has joined #openstack-neutron | 06:40 | |
| *** enikanorov2 has joined #openstack-neutron | 06:41 | |
| *** shashankhegde has quit IRC | 06:41 | |
| *** yamamoto_ has joined #openstack-neutron | 06:45 | |
| *** anilvenkata has left #openstack-neutron | 06:46 | |
| *** SridharG has joined #openstack-neutron | 06:46 | |
| *** [1]evgenyf has joined #openstack-neutron | 06:47 | |
| *** red_trela has quit IRC | 06:48 | |
| *** yamamoto has quit IRC | 06:48 | |
| *** park_heijlong has quit IRC | 06:48 | |
| *** yasu_ has quit IRC | 06:56 | |
| *** yasu_ has joined #openstack-neutron | 06:56 | |
| *** safchain has joined #openstack-neutron | 07:00 | |
| *** park_heijlong has joined #openstack-neutron | 07:01 | |
| *** Marga_ has joined #openstack-neutron | 07:01 | |
| *** Marga_ has quit IRC | 07:04 | |
| *** subscope has quit IRC | 07:05 | |
| *** luqas has joined #openstack-neutron | 07:05 | |
| *** moha_hunt has joined #openstack-neutron | 07:07 | |
| *** luqas has quit IRC | 07:10 | |
| *** fedexo has quit IRC | 07:14 | |
| *** achanda has quit IRC | 07:15 | |
| *** moshele has joined #openstack-neutron | 07:16 | |
| *** kawa2014 has joined #openstack-neutron | 07:17 | |
| *** tflynn_ has quit IRC | 07:17 | |
| *** aranjan has quit IRC | 07:19 | |
| *** Marga_ has joined #openstack-neutron | 07:19 | |
| *** Marga_ has quit IRC | 07:19 | |
| *** numan has quit IRC | 07:20 | |
| *** matrohon has joined #openstack-neutron | 07:22 | |
| *** nyechiel has joined #openstack-neutron | 07:27 | |
| *** jpena has joined #openstack-neutron | 07:27 | |
| *** criptoverso has joined #openstack-neutron | 07:27 | |
| *** sdake__ has quit IRC | 07:28 | |
| *** numan has joined #openstack-neutron | 07:28 | |
| *** anilvenkata has joined #openstack-neutron | 07:30 | |
| *** yamamoto_ has quit IRC | 07:33 | |
| *** luqas has joined #openstack-neutron | 07:33 | |
| *** yfried_ is now known as yfried|afk | 07:36 | |
| *** imcsk8 has joined #openstack-neutron | 07:38 | |
| *** luqas has quit IRC | 07:38 | |
| *** yfried|afk is now known as yfried_ | 07:38 | |
| *** jlibosva has joined #openstack-neutron | 07:42 | |
| *** tobe has joined #openstack-neutron | 07:42 | |
| *** apuimedo has joined #openstack-neutron | 07:43 | |
| *** ihrachyshka has joined #openstack-neutron | 07:43 | |
| *** achanda has joined #openstack-neutron | 07:45 | |
| *** ivar-laz_ has quit IRC | 07:45 | |
| *** mpsairam has joined #openstack-neutron | 07:45 | |
| *** ivar-lazzaro has joined #openstack-neutron | 07:46 | |
| *** yfried_ has quit IRC | 07:46 | |
| *** jlanoux has joined #openstack-neutron | 07:46 | |
| *** roeyc has joined #openstack-neutron | 07:49 | |
| *** pece has joined #openstack-neutron | 07:50 | |
| *** yamamoto has joined #openstack-neutron | 07:54 | |
| *** fifieldt has joined #openstack-neutron | 07:57 | |
| *** lukasa has joined #openstack-neutron | 07:58 | |
| *** fifieldt has quit IRC | 07:58 | |
| *** jistr has joined #openstack-neutron | 07:58 | |
| *** park_heijlong has quit IRC | 07:59 | |
| *** achanda has quit IRC | 08:01 | |
| *** devvesa has joined #openstack-neutron | 08:03 | |
| *** moha_hunt has quit IRC | 08:04 | |
| *** lukasa has quit IRC | 08:05 | |
| *** roeyc has quit IRC | 08:06 | |
| *** ygbo has joined #openstack-neutron | 08:12 | |
| *** red_trela has joined #openstack-neutron | 08:12 | |
| openstackgerrit | LIU Yulong proposed openstack/python-neutronclient: Add missing tenant_id to lbaas-v2 resources creation https://review.openstack.org/181822 | 08:15 |
| *** jlibosva has quit IRC | 08:20 | |
| *** jlibosva has joined #openstack-neutron | 08:20 | |
| *** redbeard has joined #openstack-neutron | 08:22 | |
| *** yfried has joined #openstack-neutron | 08:24 | |
| *** yog__ has quit IRC | 08:28 | |
| *** rossella_s has quit IRC | 08:28 | |
| *** ihrachyshka has quit IRC | 08:30 | |
| *** dguitarbite has joined #openstack-neutron | 08:32 | |
| *** lori|away is now known as lori | 08:33 | |
| *** sudipto has quit IRC | 08:34 | |
| *** ronis has joined #openstack-neutron | 08:36 | |
| *** zhhuabj_ has quit IRC | 08:38 | |
| openstackgerrit | Moshe Levi proposed openstack/neutron: mlnx MD: mlnx_direct removal https://review.openstack.org/181672 | 08:38 |
| *** sudipto has joined #openstack-neutron | 08:39 | |
| *** sbalukoff has quit IRC | 08:39 | |
| openstackgerrit | ShaoRan Xiao proposed openstack/neutron: [WIP]Definition of the Virtual Network Performance interface https://review.openstack.org/164814 | 08:40 |
| *** rossella_s has joined #openstack-neutron | 08:41 | |
| *** roeyc has joined #openstack-neutron | 08:41 | |
| *** ajo has joined #openstack-neutron | 08:42 | |
| *** sbalukoff has joined #openstack-neutron | 08:43 | |
| *** yamamoto has quit IRC | 08:45 | |
| *** achanda has joined #openstack-neutron | 08:45 | |
| *** yog__ has joined #openstack-neutron | 08:45 | |
| *** aranjan has joined #openstack-neutron | 08:50 | |
| *** zhhuabj_ has joined #openstack-neutron | 08:50 | |
| *** fawadkhaliq has joined #openstack-neutron | 08:52 | |
| *** gildub has joined #openstack-neutron | 08:54 | |
| *** aranjan has quit IRC | 08:55 | |
| *** erkules has quit IRC | 08:55 | |
| *** erkules has joined #openstack-neutron | 08:56 | |
| *** park_heijlong has joined #openstack-neutron | 09:04 | |
| *** park_heijlong has quit IRC | 09:09 | |
| *** hichihara has quit IRC | 09:09 | |
| *** yog__ has quit IRC | 09:12 | |
| *** dtynan has left #openstack-neutron | 09:19 | |
| *** aix has joined #openstack-neutron | 09:20 | |
| *** dtynan has joined #openstack-neutron | 09:22 | |
| *** yog__ has joined #openstack-neutron | 09:23 | |
| *** yamamoto has joined #openstack-neutron | 09:24 | |
| *** yfauser has joined #openstack-neutron | 09:24 | |
| *** yfauser has left #openstack-neutron | 09:24 | |
| *** aswadr has joined #openstack-neutron | 09:33 | |
| openstackgerrit | Sridhar Gaddam proposed openstack/neutron: Support BP:ipv6-router in Neutron HA Router https://review.openstack.org/173206 | 09:34 |
| *** zhhuabj_ has quit IRC | 09:35 | |
| openstackgerrit | venkata anil proposed openstack/neutron-vpnaas: Assign external_ip based on ip version of peer_address https://review.openstack.org/181842 | 09:36 |
| *** red_trela has quit IRC | 09:40 | |
| *** achanda has quit IRC | 09:40 | |
| openstackgerrit | Cyril Roelandt proposed openstack/neutron: Allow users to run 'tox -epy34' https://review.openstack.org/181277 | 09:41 |
| *** ZZelle has joined #openstack-neutron | 09:42 | |
| *** subscope has joined #openstack-neutron | 09:43 | |
| *** zhhuabj_ has joined #openstack-neutron | 09:47 | |
| openstackgerrit | ShaoRan Xiao proposed openstack/neutron: [WIP]Definition of the Virtual Network Performance interface https://review.openstack.org/164814 | 09:52 |
| *** fawadkhaliq has quit IRC | 09:56 | |
| *** roeyc has quit IRC | 09:58 | |
| *** merooney has joined #openstack-neutron | 09:58 | |
| *** ihrachyshka has joined #openstack-neutron | 10:00 | |
| *** dimsum__ has joined #openstack-neutron | 10:02 | |
| *** yfauser has joined #openstack-neutron | 10:03 | |
| *** yfauser has left #openstack-neutron | 10:03 | |
| *** red_trela has joined #openstack-neutron | 10:09 | |
| *** red_trela has quit IRC | 10:11 | |
| *** pc_m has joined #openstack-neutron | 10:13 | |
| *** moha_hunt has joined #openstack-neutron | 10:15 | |
| *** nmagnezi has quit IRC | 10:19 | |
| *** nmagnezi has joined #openstack-neutron | 10:22 | |
| *** dguitarbite has quit IRC | 10:24 | |
| *** dguitarbite has joined #openstack-neutron | 10:27 | |
| pece | Hello, does anuone know if 'shared' attribute in metering labels is supposed to work in Juno release? | 10:27 |
| *** jlibosva has quit IRC | 10:28 | |
| *** andyhky has quit IRC | 10:29 | |
| *** jlibosva has joined #openstack-neutron | 10:30 | |
| *** andyhky has joined #openstack-neutron | 10:31 | |
| *** fawadkhaliq has joined #openstack-neutron | 10:31 | |
| *** yuanying has quit IRC | 10:34 | |
| openstackgerrit | John Davidge proposed openstack/neutron: Implement Neutron IPv6 Prefix Delegation https://review.openstack.org/158697 | 10:36 |
| *** baoli has joined #openstack-neutron | 10:37 | |
| *** lukasa has joined #openstack-neutron | 10:39 | |
| *** aranjan has joined #openstack-neutron | 10:39 | |
| *** baoli has quit IRC | 10:41 | |
| *** aranjan has quit IRC | 10:43 | |
| *** [1]evgenyf has quit IRC | 10:43 | |
| *** lukasa has quit IRC | 10:47 | |
| *** coolsvap is now known as coolsvap|afk | 10:48 | |
| *** evgenyf has joined #openstack-neutron | 10:54 | |
| *** yalie has quit IRC | 10:54 | |
| openstackgerrit | Salvatore Orlando proposed openstack/neutron: Deprecate quota_items, register resources upon REST initialization https://review.openstack.org/181593 | 11:04 |
| *** roeyc has joined #openstack-neutron | 11:04 | |
| *** moha_hunt has quit IRC | 11:05 | |
| *** jlibosva has quit IRC | 11:07 | |
| *** jlibosva has joined #openstack-neutron | 11:08 | |
| *** zhenguo has joined #openstack-neutron | 11:10 | |
| *** lihit has joined #openstack-neutron | 11:20 | |
| *** mpsairam has quit IRC | 11:22 | |
| *** simon-AS559 has quit IRC | 11:28 | |
| *** EmilienM|afk is now known as EmilienM | 11:28 | |
| anilvenkata | pc_m Hi paul Michali, can you please review my patch for external ip when both ipv4 and ipv6 gateway available https://review.openstack.org/181842 | 11:28 |
| anilvenkata | pc_m: I put it in WIP, If you have any suggestions, please tell me, I still have to work on unit tests | 11:29 |
| *** lihit has quit IRC | 11:30 | |
| *** simon-AS559 has joined #openstack-neutron | 11:30 | |
| *** simon-AS5591 has joined #openstack-neutron | 11:32 | |
| *** jaypipes has joined #openstack-neutron | 11:32 | |
| *** simon-AS559 has quit IRC | 11:34 | |
| *** trinaths has quit IRC | 11:35 | |
| *** SridharG has left #openstack-neutron | 11:42 | |
| *** subscope has quit IRC | 11:44 | |
| *** subscope has joined #openstack-neutron | 11:47 | |
| openstackgerrit | shihanzhang proposed openstack/neutron: Allow updating port 'binding:host_id' be None https://review.openstack.org/181867 | 11:48 |
| *** jpena is now known as jpena|lunch | 11:52 | |
| *** nuritv has joined #openstack-neutron | 11:53 | |
| *** tobe has quit IRC | 11:53 | |
| *** baoli has joined #openstack-neutron | 11:56 | |
| *** claudiub has joined #openstack-neutron | 11:58 | |
| *** dimsum__ has quit IRC | 12:03 | |
| *** boris-42 has joined #openstack-neutron | 12:04 | |
| *** luqas has joined #openstack-neutron | 12:06 | |
| *** mwagner_lap has quit IRC | 12:07 | |
| *** jgalvin201514 has quit IRC | 12:09 | |
| *** dave-mccowan has joined #openstack-neutron | 12:12 | |
| *** Marga_ has joined #openstack-neutron | 12:12 | |
| *** claudiub has quit IRC | 12:13 | |
| *** _dguerri is now known as dguerri | 12:13 | |
| *** atuvenie has joined #openstack-neutron | 12:15 | |
| *** claudiub has joined #openstack-neutron | 12:16 | |
| *** simon-AS5591 has quit IRC | 12:16 | |
| *** raildo has joined #openstack-neutron | 12:19 | |
| *** Marga_ has quit IRC | 12:21 | |
| *** Marga_ has joined #openstack-neutron | 12:22 | |
| *** banix has joined #openstack-neutron | 12:22 | |
| *** fawadkhaliq has quit IRC | 12:26 | |
| *** tkatarki has joined #openstack-neutron | 12:26 | |
| jaypipes | ZZelle: around? I'm looking at your IPAM subnet pool PGsql lock fix... | 12:27 |
| *** aranjan has joined #openstack-neutron | 12:28 | |
| *** sudipto has quit IRC | 12:30 | |
| *** yalie has joined #openstack-neutron | 12:31 | |
| *** jschwarz__ has joined #openstack-neutron | 12:31 | |
| *** banix has quit IRC | 12:31 | |
| *** yfried is now known as yfried|afk | 12:32 | |
| *** roeyc has quit IRC | 12:32 | |
| *** aranjan has quit IRC | 12:33 | |
| jaypipes | ZZelle: what I'm wondering is how the quota reservations are done in this module... I don't see any reserve/commit-or-rollback-reservation strategy used. I only see a check of quotas, but not where the quotas are updated. Could you advise where I can find where the quotas are actually updated? | 12:34 |
| *** vikram has joined #openstack-neutron | 12:35 | |
| *** vikram has quit IRC | 12:38 | |
| *** sdake has joined #openstack-neutron | 12:39 | |
| openstackgerrit | Ihar Hrachyshka proposed openstack/neutron: policy: cleanup deprecation code to handle old extension:xxx rules https://review.openstack.org/176710 | 12:40 |
| *** sdake_ has joined #openstack-neutron | 12:41 | |
| *** fawadkhaliq has joined #openstack-neutron | 12:44 | |
| *** sdake has quit IRC | 12:44 | |
| *** cference has joined #openstack-neutron | 12:44 | |
| *** roeyc has joined #openstack-neutron | 12:48 | |
| *** Longgeek_ has quit IRC | 12:48 | |
| *** Longgeek has joined #openstack-neutron | 12:49 | |
| *** fawadk has joined #openstack-neutron | 12:50 | |
| *** mwagner_lap has joined #openstack-neutron | 12:50 | |
| *** kawa2014 has quit IRC | 12:50 | |
| *** fawadkhaliq has quit IRC | 12:50 | |
| *** kawa2014 has joined #openstack-neutron | 12:51 | |
| *** yatin has quit IRC | 12:51 | |
| *** sdake_ is now known as sdake | 12:55 | |
| *** inara has quit IRC | 12:56 | |
| *** inara has joined #openstack-neutron | 12:56 | |
| *** gildub has quit IRC | 12:56 | |
| pc_m | anilvenkata: hi. I'll look at it in a few mins. | 12:59 |
| anilvenkata | pc_m thanks Pc_m | 13:00 |
| *** merooney has quit IRC | 13:03 | |
| *** jpena|lunch is now known as jpena | 13:07 | |
| *** fawadkhaliq has joined #openstack-neutron | 13:08 | |
| *** merooney has joined #openstack-neutron | 13:09 | |
| *** fawadk has quit IRC | 13:11 | |
| openstackgerrit | Rossella Sblendido proposed openstack/neutron: L2 agent RPC add new RPC calls https://review.openstack.org/147435 | 13:12 |
| *** ekuris has quit IRC | 13:14 | |
| openstackgerrit | Sam Betts proposed openstack/neutron: Make l3 agent prefix delegation pluggable https://review.openstack.org/176272 | 13:15 |
| openstackgerrit | Ian Wienand proposed openstack/neutron: Ensure netfilter is enabled for bridges https://review.openstack.org/180867 | 13:15 |
| *** Longgeek has quit IRC | 13:15 | |
| *** Longgeek has joined #openstack-neutron | 13:16 | |
| *** ekuris has joined #openstack-neutron | 13:17 | |
| *** fifieldt has joined #openstack-neutron | 13:17 | |
| *** fifieldt has quit IRC | 13:17 | |
| *** TrevorV|Home has joined #openstack-neutron | 13:17 | |
| *** blues-man has joined #openstack-neutron | 13:18 | |
| *** rpothier has joined #openstack-neutron | 13:20 | |
| *** josecastroleon has joined #openstack-neutron | 13:21 | |
| *** jckasper has quit IRC | 13:21 | |
| *** chlong has joined #openstack-neutron | 13:22 | |
| *** jckasper has joined #openstack-neutron | 13:22 | |
| *** tiswanso has joined #openstack-neutron | 13:25 | |
| *** lukasa has joined #openstack-neutron | 13:25 | |
| *** kawa2014 has quit IRC | 13:28 | |
| *** aranjan has joined #openstack-neutron | 13:29 | |
| *** fawadkhaliq has quit IRC | 13:29 | |
| *** baoli has quit IRC | 13:30 | |
| *** baoli_ has joined #openstack-neutron | 13:31 | |
| *** jgrimm is now known as zz_jgrimm | 13:32 | |
| *** aranjan has quit IRC | 13:34 | |
| *** b3nt_pin has quit IRC | 13:36 | |
| *** lukasa has quit IRC | 13:37 | |
| *** beagles has quit IRC | 13:38 | |
| *** rushil has joined #openstack-neutron | 13:38 | |
| *** b3nt_pin has joined #openstack-neutron | 13:38 | |
| *** b3nt_pin is now known as beagles | 13:38 | |
| *** yamamoto has quit IRC | 13:39 | |
| Murali | hi | 13:39 |
| *** yamamoto has joined #openstack-neutron | 13:39 | |
| Murali | I am seeing the "failed 2/20: up 13.62. request failed" when I login to vm | 13:39 |
| *** dimsum__ has joined #openstack-neutron | 13:39 | |
| *** prad has joined #openstack-neutron | 13:39 | |
| Murali | and I am able to get the metadata in the vm | 13:39 |
| *** ivar-lazzaro has quit IRC | 13:41 | |
| *** kawa2014 has joined #openstack-neutron | 13:41 | |
| Murali | can some body to get my metadata in Vm | 13:41 |
| Murali | http://pastebin.com/WS2SFqvL | 13:41 |
| *** ivar-lazzaro has joined #openstack-neutron | 13:41 | |
| *** julim has joined #openstack-neutron | 13:42 | |
| *** kawa2014 has quit IRC | 13:43 | |
| *** nihilifer has joined #openstack-neutron | 13:43 | |
| *** kawa2014 has joined #openstack-neutron | 13:43 | |
| *** emagana has quit IRC | 13:43 | |
| *** merooney has quit IRC | 13:44 | |
| *** yamamoto has quit IRC | 13:44 | |
| *** yfried|afk has quit IRC | 13:44 | |
| *** yfried|afk has joined #openstack-neutron | 13:45 | |
| *** ekuris has quit IRC | 13:46 | |
| *** rushil has quit IRC | 13:46 | |
| *** ekuris has joined #openstack-neutron | 13:46 | |
| *** prad has quit IRC | 13:47 | |
| *** yog__ has quit IRC | 13:47 | |
| Murali | does anybody had faced this issue. pls share me your insight | 13:48 |
| *** prad has joined #openstack-neutron | 13:48 | |
| *** amuller has joined #openstack-neutron | 13:49 | |
| *** shwetaap has joined #openstack-neutron | 13:52 | |
| *** dimsum__ has quit IRC | 13:53 | |
| *** apuimedo has quit IRC | 13:53 | |
| *** marrusl has joined #openstack-neutron | 13:53 | |
| *** markvoelker has quit IRC | 13:56 | |
| *** joshw42 has joined #openstack-neutron | 13:57 | |
| *** Longgeek has quit IRC | 14:00 | |
| *** Longgeek has joined #openstack-neutron | 14:01 | |
| *** merooney has joined #openstack-neutron | 14:01 | |
| *** wznoinsk has joined #openstack-neutron | 14:02 | |
| *** smccully has quit IRC | 14:02 | |
| *** smccully has joined #openstack-neutron | 14:02 | |
| *** yamamoto has joined #openstack-neutron | 14:02 | |
| *** Longgeek_ has joined #openstack-neutron | 14:03 | |
| *** Longgeek has quit IRC | 14:04 | |
| *** rushil has joined #openstack-neutron | 14:04 | |
| nuritv | Hi, i am looking for reviews for this small fix :https://review.openstack.org/#/c/162648/. was suppose to be part of Kilo but postponed due to CF | 14:04 |
| *** joshw42 has quit IRC | 14:05 | |
| openstackgerrit | Jakub Libosvar proposed openstack/neutron: Remove H305 from pep8 check https://review.openstack.org/181904 | 14:06 |
| anilvenkata | pc_m: Hi Paul Michali, please submit your patch for https://bugs.launchpad.net/neutron/+bug/1440655 , one more duplicate bug( https://bugs.launchpad.net/neutron/+bug/1453656) is raised for the same issue :) | 14:06 |
| openstack | Launchpad bug 1440655 in neutron "VPNaas:service ipsec status gives incorrect status of the ipsec tunnel." [Undecided,Confirmed] - Assigned to Aniruddha Singh Gautam (aniruddha-gautam) | 14:06 |
| openstack | Launchpad bug 1440655 in neutron "duplicate for #1453656 VPNaas:service ipsec status gives incorrect status of the ipsec tunnel." [Undecided,Confirmed] - Assigned to Aniruddha Singh Gautam (aniruddha-gautam) | 14:06 |
| *** openstackgerrit has quit IRC | 14:06 | |
| *** openstackgerrit has joined #openstack-neutron | 14:07 | |
| pc_m | anilvenkata: Someone assigned themselves to 1440655. I commented, asking if they are working on it and if not, I'd fix. No response, so I think I'll just reassign and fix it. | 14:08 |
| *** rossella_s has quit IRC | 14:09 | |
| *** kbringard has joined #openstack-neutron | 14:09 | |
| *** rook has quit IRC | 14:09 | |
| *** rossella_s has joined #openstack-neutron | 14:09 | |
| openstackgerrit | Cyril Roelandt proposed openstack/neutron: Allow users to run 'tox -epy34' https://review.openstack.org/181277 | 14:10 |
| *** neeti has quit IRC | 14:10 | |
| *** rook has joined #openstack-neutron | 14:11 | |
| *** anilvenkata has quit IRC | 14:11 | |
| *** yamamoto has quit IRC | 14:11 | |
| *** dboik has joined #openstack-neutron | 14:11 | |
| *** numan has quit IRC | 14:11 | |
| *** dave-mccowan has quit IRC | 14:14 | |
| *** sdake_ has joined #openstack-neutron | 14:16 | |
| *** dims has joined #openstack-neutron | 14:16 | |
| *** ferest has joined #openstack-neutron | 14:16 | |
| *** jobewan has joined #openstack-neutron | 14:16 | |
| *** zzzeek has joined #openstack-neutron | 14:18 | |
| *** openstack has joined #openstack-neutron | 14:20 | |
| *** chem has joined #openstack-neutron | 14:20 | |
| *** yasu_ has quit IRC | 14:20 | |
| *** ferest has quit IRC | 14:21 | |
| *** jobewan has joined #openstack-neutron | 14:21 | |
| *** jobewan has quit IRC | 14:21 | |
| *** jobewan has joined #openstack-neutron | 14:22 | |
| *** jobewan has quit IRC | 14:22 | |
| *** yalie has quit IRC | 14:22 | |
| *** jobewan has joined #openstack-neutron | 14:22 | |
| *** irenab has quit IRC | 14:23 | |
| *** numan has joined #openstack-neutron | 14:25 | |
| *** ArchiFleKs has joined #openstack-neutron | 14:25 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: Move full-stack logs post-tests https://review.openstack.org/180882 | 14:25 |
| jschwarz__ | pc_m, ^ please have another look at make sure it suits your needs :) | 14:25 |
| *** lukasa has joined #openstack-neutron | 14:26 | |
| *** jschwarz__ is now known as jschwarz | 14:26 | |
| pc_m | jschwarz: ok | 14:26 |
| jschwarz | ty | 14:26 |
| sc68cal | HenryG: hey, sorry to bug you about this, but I think that technical brief on Cisco's website that is linked from https://wiki.openstack.org/wiki/NEUTRON-IPV6-MANUAL has errors | 14:28 |
| *** breitz has joined #openstack-neutron | 14:28 | |
| *** anilvenkata has joined #openstack-neutron | 14:28 | |
| *** dave-mccowan has joined #openstack-neutron | 14:29 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: Support for concurrent full-stack tests https://review.openstack.org/180972 | 14:31 |
| *** sdake has joined #openstack-neutron | 14:32 | |
| jschwarz | nmagnezi, ping | 14:32 |
| jschwarz | nmagnezi, can you please review https://review.openstack.org/#/c/174268/ and https://review.openstack.org/#/c/174384/ for me? :) | 14:32 |
| nmagnezi | jschwarz, will do :-) | 14:33 |
| *** neophy has joined #openstack-neutron | 14:33 | |
| pc_m | jschwarz: So what's the decision on the path for logs? They initially were hard coded to /opt/stack/, which won't be right for gate where it is /opt/stack/new/. Then, there was some mention of /tmp having more files than just the logs (or was the intent to use some job name specific area in /tmp)? | 14:33 |
| *** vikram has joined #openstack-neutron | 14:34 | |
| pc_m | jschwarz: I see it's filtering by .log files. What if > 1 job running? | 14:34 |
| *** lukasa has quit IRC | 14:34 | |
| * pc_m or are we being too anal on this? | 14:34 | |
| *** lukasa has joined #openstack-neutron | 14:35 | |
| *** sdake_ has quit IRC | 14:35 | |
| *** lukasa has quit IRC | 14:36 | |
| *** prad has quit IRC | 14:37 | |
| *** zz_jgrimm is now known as jgrimm | 14:37 | |
| *** lukasa has joined #openstack-neutron | 14:37 | |
| *** prad has joined #openstack-neutron | 14:37 | |
| *** jprovazn has joined #openstack-neutron | 14:38 | |
| *** mestery has quit IRC | 14:38 | |
| *** neophy has quit IRC | 14:40 | |
| *** annegentle has joined #openstack-neutron | 14:40 | |
| *** Murali has quit IRC | 14:40 | |
| *** emagana has joined #openstack-neutron | 14:42 | |
| jschwarz | pc_m, the logs directory (and the directory from which the post_hook_test.sh and infra take the logs from) is /opt/stack/logs | 14:42 |
| *** itzikb has quit IRC | 14:42 | |
| jschwarz | and that's where all the log files should be if you want them to be uploaded | 14:42 |
| jschwarz | obviously /tmp holds more than just files, but I'm fine putting full-stack logs there since most other projects put logs in /var/log or /var/run or something similar, so we can just take /tmp/*.log, gzip them and move them to the log dir /opt/stack/logs | 14:43 |
| jschwarz | Also, 2 jobs = 2 different VMs, so they don't run on the same machine and don't share the /tmp folder anyway. | 14:44 |
| pc_m | jschwarz: OK. So no problem of picking up logs form other jobs. | 14:44 |
| jschwarz | pc_m, yep | 14:44 |
| pc_m | jschwarz: thanks | 14:44 |
| jschwarz | pc_m, thank you :) | 14:44 |
| jschwarz | nmagnezi, thanks | 14:45 |
| *** amotoki_ has joined #openstack-neutron | 14:45 | |
| *** anteaya has joined #openstack-neutron | 14:45 | |
| pc_m | jschwarz: +1'ed (would +2, but I can't :) | 14:46 |
| jschwarz | :D | 14:46 |
| openstackgerrit | Sam Betts proposed openstack/neutron: Add an internal DHCPv6 client for Prefix Delegation https://review.openstack.org/161381 | 14:47 |
| *** banix has joined #openstack-neutron | 14:52 | |
| *** saggi has quit IRC | 14:52 | |
| ZZelle | jaypipes, hi | 14:54 |
| *** ekuris has quit IRC | 14:54 | |
| ZZelle | jaypipes, by quotas, you mean used/left resources in quotas or quota limits? | 14:56 |
| *** nmagnezi has quit IRC | 14:58 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: Support for concurrent full-stack tests https://review.openstack.org/180972 | 14:58 |
| HenryG | sc68cal: no need to apologize, feedback is appreciated! What errors do you see? | 14:59 |
| ZZelle | jschwarz, ^, thx | 14:59 |
| jschwarz | ZZelle, :) | 14:59 |
| amuller | jschwarz: about https://review.openstack.org/#/c/180882/3 | 14:59 |
| jschwarz | ZZelle, do you have anything need reviewing? | 14:59 |
| ZZelle | jschwarz, ah you forgot my comment :) | 15:00 |
| amuller | jschwarz: how do I debug local issues now? | 15:00 |
| jschwarz | amuller, logs will be saved to /tmp instead of /opt/stack/logs | 15:00 |
| jschwarz | that's it :) | 15:00 |
| sc68cal | HenryG: the main thing is around some ipv6_ra_mode and ipv6_address_mode combos | 15:00 |
| jschwarz | ZZelle, what did I forget? | 15:00 |
| sc68cal | HenryG: SridharG also spotted some stuff in the gerrit review, where I took the content from the wiki and added it to the networking guide (https://review.openstack.org//#/c/181049/2/doc/networking-guide/source/ipv6.rst) | 15:01 |
| jschwarz | ZZelle, ahh I didn't see that :< https://review.openstack.org/#/c/180972/4..5/neutron/tests/fullstack/fullstack_fixtures.py | 15:02 |
| *** jpena has quit IRC | 15:02 | |
| *** tiswanso has quit IRC | 15:02 | |
| *** markvoelker has joined #openstack-neutron | 15:03 | |
| HenryG | sc68cal: The mode attributes bite us again :( | 15:04 |
| *** wwriverrat has joined #openstack-neutron | 15:04 | |
| openstackgerrit | Miguel Angel Ajo proposed openstack/neutron-specs: Neutron QoS API Extension https://review.openstack.org/88599 | 15:04 |
| *** dboik has quit IRC | 15:05 | |
| *** jpena has joined #openstack-neutron | 15:05 | |
| sc68cal | HenryG: yeah..... :( | 15:05 |
| jschwarz | ZZelle, done | 15:05 |
| *** jckasper has quit IRC | 15:05 | |
| sc68cal | HenryG: the problem is even the specs we filed for the ipv6 attrs didn't agree | 15:05 |
| sc68cal | HenryG: between my spec and baoli_ 's | 15:06 |
| openstackgerrit | John Schwarz proposed openstack/neutron: Support for concurrent full-stack tests https://review.openstack.org/180972 | 15:06 |
| *** daneyon_ has joined #openstack-neutron | 15:06 | |
| sc68cal | and i'm trying to put the toothpaste back in the tube way late | 15:06 |
| sc68cal | it's my fault, at the end of the day | 15:06 |
| *** jckasper has joined #openstack-neutron | 15:07 | |
| HenryG | sc68cal: No it's not your fault | 15:07 |
| *** rook has quit IRC | 15:07 | |
| HenryG | sc68cal: It's the reality of getting ipv6 support into the existing L3 code | 15:08 |
| *** tidwellr has joined #openstack-neutron | 15:08 | |
| HenryG | With multiple developers working on different parts of it | 15:09 |
| *** daneyon has quit IRC | 15:09 | |
| HenryG | sc68cal: I think it makes most sense to converge on the doc review first, then I'll make sure the Cisco posting is updated | 15:09 |
| HenryG | The doc should match what the code does today | 15:10 |
| HenryG | Which is not always easy to figure out for every corner case | 15:10 |
| sc68cal | yeah | 15:11 |
| *** dims has quit IRC | 15:12 | |
| *** aix has quit IRC | 15:12 | |
| *** aswadr has quit IRC | 15:14 | |
| *** yamahata has joined #openstack-neutron | 15:14 | |
| *** wwriverrat has left #openstack-neutron | 15:16 | |
| *** baoli_ has quit IRC | 15:16 | |
| *** aranjan has joined #openstack-neutron | 15:18 | |
| jschwarz | amuller, gah, damn logs | 15:19 |
| *** roeyc has quit IRC | 15:19 | |
| jschwarz | amuller, turns out 180882 isn't rebased on carl's patch, so fullstack isn't enabled there | 15:19 |
| jschwarz | and the static "mv all logs" fail because no such files exists, so the test fails | 15:19 |
| *** dannywilson has joined #openstack-neutron | 15:20 | |
| *** roeyc has joined #openstack-neutron | 15:20 | |
| *** rook has joined #openstack-neutron | 15:20 | |
| *** vthapar has quit IRC | 15:21 | |
| *** fifieldt has joined #openstack-neutron | 15:22 | |
| *** fifieldt has quit IRC | 15:23 | |
| *** aranjan has quit IRC | 15:23 | |
| *** neophy has joined #openstack-neutron | 15:27 | |
| *** ronis has quit IRC | 15:28 | |
| *** atuvenie has quit IRC | 15:29 | |
| *** sbalukoff has quit IRC | 15:30 | |
| *** ivar-lazzaro has quit IRC | 15:32 | |
| kevinbenton | salv-orlando: ping | 15:32 |
| *** ivar-lazzaro has joined #openstack-neutron | 15:33 | |
| *** armax has joined #openstack-neutron | 15:33 | |
| *** ygbo has quit IRC | 15:34 | |
| *** scheuran has quit IRC | 15:34 | |
| salv-orlando | aloha kevinbenton | 15:34 |
| salv-orlando | how can I help you today, sir? | 15:35 |
| *** tiswanso has joined #openstack-neutron | 15:35 | |
| *** dboik has joined #openstack-neutron | 15:35 | |
| *** jbell8 has joined #openstack-neutron | 15:35 | |
| kevinbenton | i was looking at https://review.openstack.org/#/c/181867/ | 15:36 |
| kevinbenton | with extensions like portbinding, are they not prepopulated when the API caller doesn't provide them? | 15:36 |
| kevinbenton | salv-orlando: ^^ | 15:36 |
| *** annegentle has quit IRC | 15:37 | |
| salv-orlando | kevinbenton: do you mean prepopulated as "default values are automatically set"? | 15:37 |
| salv-orlando | or just that the attribute is there | 15:37 |
| salv-orlando | in the dict, I mean | 15:37 |
| kevinbenton | just that the key is in the dict | 15:37 |
| kevinbenton | yeah | 15:37 |
| *** annegentle has joined #openstack-neutron | 15:38 | |
| *** atuvenie has joined #openstack-neutron | 15:38 | |
| salv-orlando | perhaps that's right. the current code as it is won't work. He needs an explicit check for None | 15:38 |
| *** yamamoto has joined #openstack-neutron | 15:39 | |
| *** Swami has joined #openstack-neutron | 15:40 | |
| kevinbenton | salv-orlando: or was it that the update operation didn't populate the whole dict? | 15:40 |
| *** jbell8 has quit IRC | 15:40 | |
| *** daneyon_ has quit IRC | 15:41 | |
| salv-orlando | kevinbenton: the update operation does not populate the whole thing because it implements patch semantic, but what you pointed out made me realize I forgot to verify whether that method is called only on update | 15:41 |
| kevinbenton | salv-orlando: it's called both times | 15:41 |
| *** daneyon has joined #openstack-neutron | 15:41 | |
| kevinbenton | salv-orlando: but i don't think it really harms the create case | 15:41 |
| kevinbenton | salv-orlando: so this logic should be okay | 15:42 |
| *** rushil has quit IRC | 15:42 | |
| *** dboik_ has joined #openstack-neutron | 15:42 | |
| *** britthou_ has quit IRC | 15:43 | |
| salv-orlando | kevinbenton: that's probably correct. On create the effect would be the same as before is host_id is omitted | 15:43 |
| *** mchestr has joined #openstack-neutron | 15:43 | |
| *** atuvenie has quit IRC | 15:43 | |
| *** dboik has quit IRC | 15:44 | |
| *** vikram has quit IRC | 15:45 | |
| *** rushil has joined #openstack-neutron | 15:45 | |
| *** markvoelker has quit IRC | 15:45 | |
| *** baoli has joined #openstack-neutron | 15:46 | |
| *** ArchiFleKs has quit IRC | 15:46 | |
| *** neophy has quit IRC | 15:46 | |
| *** anshul has quit IRC | 15:46 | |
| *** rwsu has joined #openstack-neutron | 15:48 | |
| kevinbenton | salv-orlando: thanks | 15:48 |
| amotoki_ | kevinbenton: salv-orlando: I think the confusion comes from the inconsistency in the current neutron API. None and empty string/list/dict are treated as same in some cases and are treated differently in some cases... | 15:48 |
| amotoki_ | we need to make our API more consistent when we design the next version of API. | 15:49 |
| *** nlahouti has joined #openstack-neutron | 15:50 | |
| kevinbenton | amotoki_: i think this is something that can be helped with just a new internal API that passes objects | 15:51 |
| *** carl_baldwin has joined #openstack-neutron | 15:52 | |
| sc68cal | amotoki_: ++++++++ | 15:52 |
| amotoki_ | kevinbenton: totally agree. previously I tried to address this inconsistency, but there are many... :-( | 15:52 |
| * sc68cal hopes he can finally get that pony he's always wanted | 15:53 | |
| *** padkrish has joined #openstack-neutron | 15:53 | |
| *** mlavalle has joined #openstack-neutron | 15:54 | |
| openstackgerrit | Jakub Libosvar proposed openstack/neutron: Remove H305 from tox.ini pep8 ignore list https://review.openstack.org/181904 | 15:56 |
| *** rushil has quit IRC | 15:58 | |
| ZZelle | carl_baldwin, , haleyb, hi | 15:58 |
| *** baoli has quit IRC | 15:59 | |
| *** annegentle has quit IRC | 15:59 | |
| haleyb | ZZelle: hi | 15:59 |
| *** blues-man has quit IRC | 16:00 | |
| ZZelle | haleyb, carl_baldwin, i am trying to disable conntrack tracking with stateless nat | 16:00 |
| *** annegentle has joined #openstack-neutron | 16:00 | |
| *** merooney has quit IRC | 16:01 | |
| ZZelle | haleyb, carl_baldwin, sudo iptables -i qg-+ -t raw -A PREROUTING -j NOTRACK works for ingress traffic | 16:02 |
| haleyb | ZZelle: like with a -j CT --notrack target ? | 16:02 |
| *** yeungp has joined #openstack-neutron | 16:02 | |
| ZZelle | haleyb, carl_baldwin, but for egress traffic i cannot do: sudo iptables -t raw -A PREROUTING -o qg-+ -j NOTRACK because -o is not available | 16:03 |
| *** banix has quit IRC | 16:03 | |
| *** _cjones_ has joined #openstack-neutron | 16:04 | |
| haleyb | ZZelle: right, perhaps POSTROUTING? or just don't use -i/-o ? That's what the DVR code does when it creates a new namespace | 16:04 |
| *** jlanoux has quit IRC | 16:05 | |
| ZZelle | haleyb, POSTROUTING in raw table? | 16:05 |
| carl_baldwin | ZZelle, I think POSTROUTING in the raw table is too late, right? | 16:05 |
| *** safchain has quit IRC | 16:05 | |
| *** _cjones_ has quit IRC | 16:05 | |
| ZZelle | carl_baldwin, i didn't check but it seems too late in netfilter workflow | 16:06 |
| *** annegentle has quit IRC | 16:06 | |
| haleyb | ZZelle: oh yeah, only PREROUTING and OUTPUT in raw table | 16:06 |
| *** _cjones_ has joined #openstack-neutron | 16:07 | |
| *** baoli has joined #openstack-neutron | 16:08 | |
| carl_baldwin | ZZelle, I guess we'd have to match on destination address. | 16:08 |
| *** rdo has quit IRC | 16:08 | |
| *** fitoduarte has quit IRC | 16:08 | |
| *** matrohon has quit IRC | 16:08 | |
| ZZelle | carl_baldwin, haleyb, yes, that's the only solution i see which implies to enable tracking on cidrs connected to the router + metadata? | 16:09 |
| *** rdo has joined #openstack-neutron | 16:10 | |
| *** merooney has joined #openstack-neutron | 16:10 | |
| *** yfried|afk is now known as yfried | 16:10 | |
| carl_baldwin | ZZelle, there sure are a lot of little gotchas, aren't there? | 16:10 |
| openstackgerrit | John Schwarz proposed openstack/neutron: Move full-stack logs post-tests https://review.openstack.org/180882 | 16:11 |
| anilvenkata | pc_m: I have replied to your review comments, can you kindly look at them | 16:11 |
| *** subscope has quit IRC | 16:12 | |
| anilvenkata | pc_m: if any changes are really required I would do them after your suggestions | 16:12 |
| *** padkrish has quit IRC | 16:12 | |
| haleyb | ZZelle: so why didn't the default notrack rule added to the FIP namespac work? | 16:13 |
| pc_m | will look at as soon as I get out of (another) meeting | 16:13 |
| jschwarz | carl_baldwin, holla. I now have 2 patches depending on https://review.openstack.org/#/c/176093/10 ... just a kind reminder ;-) | 16:13 |
| jschwarz | Also, is the gerrit not working for everyone else also? | 16:13 |
| anilvenkata | pc_m: thanks Paul | 16:13 |
| ZZelle | carl_baldwin, i am wondering if we should not just disable tracking except for metadata ... it's difficult for me to see the impacts | 16:14 |
| carl_baldwin | jschwarz, I'm putting up a new patch set now. I'm having more troubles running the functional test so I'm going to throw it up a bit blind. | 16:14 |
| *** tflynn has joined #openstack-neutron | 16:14 | |
| jschwarz | carl_baldwin, functional or full-stack? | 16:14 |
| carl_baldwin | jschwarz, I'm having intermittent trouble with gerrit today. You're not alone. | 16:15 |
| carl_baldwin | jschwarz, I meant full stack. | 16:15 |
| jschwarz | uhm | 16:15 |
| jschwarz | was afraid my patch broke the gerrit <.< | 16:16 |
| jschwarz | carl_baldwin, what kind of fullstack problems? | 16:16 |
| carl_baldwin | It doesn't seem to be creating any namespaces now. | 16:16 |
| carl_baldwin | :( | 16:16 |
| jschwarz | :O | 16:16 |
| jschwarz | carl_baldwin, I'd suggest clearing /opt/stack/logs and running the test again. does that folder has an l3-agent log file? | 16:17 |
| ZZelle | haleyb, i don't know, i have a poor knowledge of DVR | 16:17 |
| *** jistr has quit IRC | 16:17 | |
| carl_baldwin | jschwarz, Yes, I've been checking the log files in there. | 16:17 |
| openstackgerrit | Carl Baldwin proposed openstack/neutron: Append @randtoken to L3 agent namespaces in full stack tests https://review.openstack.org/176093 | 16:17 |
| carl_baldwin | ... they are getting created. | 16:17 |
| carl_baldwin | jschwarz, the patch is posted ^ | 16:17 |
| jschwarz | carl_baldwin, ty. Can you post one of those logs files to a pastebin? | 16:18 |
| carl_baldwin | jschwarz, except I forgot to remove the setup.cfg change. | 16:18 |
| * jschwarz is surprised we added a py34 check :O | 16:18 | |
| haleyb | ZZelle: i think it would happen in non-dvr mode too, but there is a '-t raw $agent-PREROUTING -j CT --notrack' rule added | 16:19 |
| openstackgerrit | Carl Baldwin proposed openstack/neutron: Append @randtoken to L3 agent namespaces in full stack tests https://review.openstack.org/176093 | 16:19 |
| jschwarz | carl_baldwin, nevermind. I'll fetch your code and run it locally. give me a few minutes | 16:19 |
| *** banix has joined #openstack-neutron | 16:19 | |
| *** jogo has quit IRC | 16:19 | |
| *** criptoverso has quit IRC | 16:19 | |
| carl_baldwin | jschwarz, Be sure to get PS12 or later. ^ | 16:19 |
| *** rdo has quit IRC | 16:19 | |
| jschwarz | yep. 2 fetches in a row ^_^ | 16:20 |
| ZZelle | haleyb, i didn't find it, what does exactly fip namespace? | 16:20 |
| haleyb | ZZelle: i will have to turn off DVR and restack to see where that gets put - oh, and should i put a patch on-top | 16:20 |
| jschwarz | carl_baldwin, it runs just fine on my machine | 16:20 |
| *** moshele has quit IRC | 16:21 | |
| *** hareeshp has joined #openstack-neutron | 16:21 | |
| ZZelle | haleyb, you can | 16:21 |
| *** devvesa has quit IRC | 16:21 | |
| *** rdo has joined #openstack-neutron | 16:21 | |
| haleyb | ZZelle: with DVR each compute node will have a floating ip (fip) namespace, but it would be easier for me to just restack and look at a real system | 16:22 |
| carl_baldwin | jschwarz, Good to hear. | 16:22 |
| *** baoli has quit IRC | 16:22 | |
| jschwarz | lets wait for the gate and see what he says | 16:22 |
| carl_baldwin | jschwarz, It always has the last work. :) | 16:22 |
| carl_baldwin | s/work/word/ | 16:22 |
| *** claudiub|2 has joined #openstack-neutron | 16:22 | |
| jschwarz | yes it does :) | 16:22 |
| jschwarz | armax, could you kindly have a final look on https://review.openstack.org/#/c/174268/ ? I want to get it merged before the summit so I can work on the follow up patch in the series during the summit (not gonna be there) | 16:23 |
| armax | jschwarz: it’s on my backlog | 16:23 |
| jschwarz | armax, excellent. thanks plenty :) | 16:24 |
| *** jlibosva has quit IRC | 16:24 | |
| *** claudiub has quit IRC | 16:25 | |
| *** sdake_ has joined #openstack-neutron | 16:25 | |
| carl_baldwin | jschwarz, You ran PS12 and it worked? Did you get a log file? | 16:25 |
| *** pcm_ has joined #openstack-neutron | 16:27 | |
| *** evgenyf has quit IRC | 16:27 | |
| *** baoli has joined #openstack-neutron | 16:28 | |
| *** padkrish has joined #openstack-neutron | 16:29 | |
| *** pc_m has quit IRC | 16:29 | |
| *** alaski has quit IRC | 16:29 | |
| *** marun has joined #openstack-neutron | 16:29 | |
| ZZelle | haleyb, iiuc, fip namespace acts as an arp gateway between external networks and router namespace so no need to track connections (L4) | 16:29 |
| *** sdake has quit IRC | 16:29 | |
| *** smccully has quit IRC | 16:29 | |
| *** jschwarz has quit IRC | 16:29 | |
| carl_baldwin | ZZelle, that is right. | 16:29 |
| *** matrohon has joined #openstack-neutron | 16:30 | |
| *** tiswanso has quit IRC | 16:30 | |
| haleyb | ZZelle: right, i was just assuming that same logic could apply everywhere (no -i/-o in rule) - i.e. in the qrouter namespace too | 16:31 |
| *** tiswanso_ has joined #openstack-neutron | 16:31 | |
| *** numan has quit IRC | 16:31 | |
| *** anilvenkata1 has joined #openstack-neutron | 16:32 | |
| *** alaski has joined #openstack-neutron | 16:35 | |
| *** Sukhdev has joined #openstack-neutron | 16:35 | |
| *** anilvenkata has quit IRC | 16:35 | |
| *** sdake has joined #openstack-neutron | 16:36 | |
| *** TrevorV|Home has quit IRC | 16:37 | |
| ZZelle | carl_baldwin, haleyb, it becomes complicated: disable tracking for natted fixed-ips (-s) when destination (-d) is no a router internal cidrs and not metadata? | 16:37 |
| *** SridharG has joined #openstack-neutron | 16:37 | |
| *** jbell8 has joined #openstack-neutron | 16:37 | |
| *** aranjan has joined #openstack-neutron | 16:38 | |
| ZZelle | source filter is required in order to support SNAT for non-natted ports, destination filter seems required if we want to allow stateful nat from internal ports | 16:39 |
| *** yamamoto has quit IRC | 16:39 | |
| *** sdake_ has quit IRC | 16:40 | |
| carl_baldwin | ZZelle, I see how this is getting complicated. I guess that is what you'd have to do. Probably have an entry matching each internal route -- including static routes -- and metadata that avoids skipping conntrack and then skip anything that doesn't hit one of those rules. | 16:40 |
| *** absubram has joined #openstack-neutron | 16:40 | |
| *** SridharG1 has joined #openstack-neutron | 16:41 | |
| *** smccully has joined #openstack-neutron | 16:42 | |
| *** jschwarz has joined #openstack-neutron | 16:42 | |
| *** Marga_ has quit IRC | 16:42 | |
| *** kawa2014 has quit IRC | 16:43 | |
| ZZelle | carl_baldwin, yes, sadly ... i am just surprised that iptables part will be more complicated than tc one :( | 16:43 |
| jschwarz | carl_baldwin, sorry, got disconnected. yes. | 16:44 |
| *** SridharG has quit IRC | 16:44 | |
| jschwarz | carl_baldwin, do you want me to upload the log file? | 16:45 |
| carl_baldwin | jschwarz, I was able to get in running again. It seems that there were lots of ovs agent and neutron server processes running. I killed them all and things got better. | 16:45 |
| *** banix has quit IRC | 16:46 | |
| ZZelle | carl_baldwin, haleyb, ttyl, it's time for me to go :) | 16:46 |
| pcm_ | anilvenkata1: hi | 16:46 |
| *** ZZelle is now known as ZZelle_ | 16:46 | |
| anilvenkata1 | pc_m yes Paul | 16:46 |
| jschwarz | carl_baldwin, sounds reasonable. https://review.openstack.org/#/c/180972/ fixes it so you might want to review that ;-) | 16:46 |
| pcm_ | anilvenkata1: RE: 181842. Have some questions... | 16:46 |
| anilvenkata1 | pc_m yes | 16:46 |
| ZZelle_ | jschwarz, it's a publicity-free channel :) | 16:47 |
| jschwarz | carl_baldwin, though the ovs agents and neutron server leftovers shouldn't be there anymore as they were also fixed... | 16:47 |
| pcm_ | anilvenkata1: You state that service drive might not have external public network connection? | 16:47 |
| *** jpena has quit IRC | 16:47 | |
| anilvenkata1 | pcm_: yes, neutron server | 16:47 |
| carl_baldwin | jschwarz, Sounds like it is getting better piece by piece. | 16:47 |
| jschwarz | ZZelle_, lol sorry, I'll wait in line ;-) | 16:47 |
| jschwarz | carl_baldwin, hopefully. I'll look into it | 16:47 |
| pcm_ | anilvenkata1: I guess I don't see how that could be true. It's running on the network node. | 16:47 |
| *** blues-man has joined #openstack-neutron | 16:48 | |
| *** ronis has joined #openstack-neutron | 16:48 | |
| *** shashankhegde has joined #openstack-neutron | 16:49 | |
| carl_baldwin | jschwarz, I saw that patch but I don't recall seeing the server leftovers get fixed. Which patch is that? | 16:49 |
| *** blues-man has quit IRC | 16:49 | |
| jschwarz | carl_baldwin, that was... lets see... https://review.openstack.org/#/c/175945/ | 16:50 |
| pcm_ | anilvenkata1: If we can, we should strive to validate in service driver, where we can prevent the dbase update and reject the command. By the time it gets to device driver, the change has been persisted. | 16:50 |
| jschwarz | carl_baldwin, but that was way in April so your code should have it | 16:50 |
| *** mlavalle has quit IRC | 16:50 | |
| pcm_ | anilvenkata1: Also, even with service driver, there can be per implementation validators, so OpenSwan can have a different validation than strongswan. | 16:51 |
| *** mlavalle has joined #openstack-neutron | 16:51 | |
| anilvenkata1 | pc_m just my guess, if neutron server is on management network, which is accessible to all other openstack services and compute nodes, and a compute node is having vpn agent which is having external network | 16:51 |
| jschwarz | carl_baldwin, how old were the leftover processes? were they launched today? yesterday? | 16:52 |
| carl_baldwin | jschwarz, Yes, my patch includes that patch. | 16:52 |
| * carl_baldwin looking at history... | 16:52 | |
| jschwarz | carl_baldwin, if you're able to reproduce the issue it will make chasing it a lot easier :) | 16:52 |
| *** dguitarbite has quit IRC | 16:52 | |
| carl_baldwin | jschwarz, May 6 | 16:52 |
| SridharG1 | can any core reviewers please review the following patch, its a clean cherry pick - https://review.openstack.org/#/c/179392/ Thank you. | 16:52 |
| *** atuvenie has joined #openstack-neutron | 16:52 | |
| jschwarz | carl_baldwin, no way you have the screen log from back then, right? | 16:53 |
| anilvenkata1 | or neutron server(q-svc) running on controller node and vpn agent running on network node and connected to external network | 16:53 |
| carl_baldwin | jschwarz, No, sorry. | 16:53 |
| *** enikanorov2 has quit IRC | 16:53 | |
| *** SridharG1 has left #openstack-neutron | 16:53 | |
| anilvenkata1 | pc_m ^ | 16:53 |
| jschwarz | carl_baldwin, I'll open a launchpad for it. Perhaps someone can reproduce this and we can solve it then | 16:54 |
| *** banix has joined #openstack-neutron | 16:54 | |
| carl_baldwin | jschwarz, Let me play around a little. I'll open a bug if I can reproduce. | 16:54 |
| jschwarz | carl_baldwin, ack, thanks. Make sure to assign it to me and I'll get to work on it soon :) | 16:54 |
| anilvenkata1 | pc_m I am just guessing if fqdn is only resolved by vpn agent, then I can't have peer address in neutron server right? | 16:55 |
| pcm_ | anilvenkata1: I thought neutron service always had access to networks. | 16:55 |
| carl_baldwin | jschwarz, it seems to clean up now. Maybe if I introduce the type of failures that I was seeing before I can tease out the bug. | 16:55 |
| *** afazekas has quit IRC | 16:55 | |
| *** padkrish has quit IRC | 16:55 | |
| jschwarz | carl_baldwin, I tried adding a 'self.fail()' to the test itself, but the processes still get cleaned | 16:56 |
| jschwarz | on the other hand, maybe the problem is in the fixture (the failure is during the start up since it can't ping the correct server) | 16:57 |
| *** pece has quit IRC | 16:58 | |
| *** yamahata has quit IRC | 16:58 | |
| *** pece has joined #openstack-neutron | 16:58 | |
| *** yamahata has joined #openstack-neutron | 16:59 | |
| *** banix has quit IRC | 16:59 | |
| anilvenkata1 | pcm_: May be you are right, we should be able to resolve fqdn on neutron server, as neutron server should be reachable for rest api | 16:59 |
| *** amotoki_ has quit IRC | 16:59 | |
| *** rushil has joined #openstack-neutron | 17:00 | |
| jschwarz | carl_baldwin, reproduces | 17:00 |
| *** pece has quit IRC | 17:00 | |
| jschwarz | carl_baldwin, added a 'return False' right after neutron/tests/fullstack/fullstack_fixtures.py#81, the fixture doesn't start and the processes aren't cleaned | 17:00 |
| pcm_ | anilvenkata1: I've been working on DevStack for so long (all in one), I forget the details of what's running where. | 17:00 |
| carl_baldwin | jschwarz, Good find. | 17:00 |
| *** matrohon has quit IRC | 17:01 | |
| *** roeyc has quit IRC | 17:01 | |
| *** pece has joined #openstack-neutron | 17:01 | |
| jschwarz | carl_baldwin, you led me to it :) | 17:01 |
| carl_baldwin | jschwarz, I faked a timeout trying to find the namespace. I got a metadata proxy that didn't get cleaned up. | 17:01 |
| carl_baldwin | ... but not the neutron server or ovs agent with mine this time. | 17:01 |
| jschwarz | carl_baldwin, part of a bigger issue that amuller raised, actually | 17:01 |
| jschwarz | carl_baldwin, the test issues a 'delete_router' but the l3 agent doesn't clean it up in time and gets killed (SIGKILL) | 17:02 |
| jschwarz | he actually sent you a mail about it last week ;-) | 17:02 |
| carl_baldwin | Really? Maybe I missed the mail. What was the subject? | 17:02 |
| carl_baldwin | jschwarz: ^ | 17:03 |
| jschwarz | carl_baldwin, "Implementing L3 agent SIGTERM handler" | 17:03 |
| anilvenkata1 | pcm_ I need your help, you said "even with service driver, there can be per implementation validators, so OpenSwan can have a different validation than strongswan.", can u kindly show where in service driver I can add validation for only openswan? | 17:03 |
| pcm_ | anilvenkata1: AFAIK, the neutron services run on the network node, so should have access to any external network. | 17:03 |
| *** Networkn3rd has joined #openstack-neutron | 17:03 | |
| pcm_ | anilvenkata1: Take a look at the cisco service driver, where there is a different validator. You could use the reference one for OpenSwan and do validation there and StrongSwan can provide a different validator, or vice versa. | 17:04 |
| carl_baldwin | jschwarz, it got in to my mail archive somehow. I have the mail. I will read it now. | 17:04 |
| *** annegentle has joined #openstack-neutron | 17:05 | |
| jschwarz | carl_baldwin, :) | 17:05 |
| *** daneyon_ has joined #openstack-neutron | 17:05 | |
| anilvenkata1 | pcm_ agree with u about external network connectivity for network node, thanks for clarifying this, I will check cisco service driver for validation | 17:05 |
| *** shashankhegde has quit IRC | 17:05 | |
| carl_baldwin | jschwarz, Just read through the bug. "I'll get to work on it once https://review.openstack.org/#/c/176093/ merges." :) Sorry that one is taking so long. | 17:06 |
| anilvenkata1 | pcm_: thanks Paul Michali for suggestions | 17:06 |
| pcm_ | anilvenkata1: Sure np. | 17:06 |
| jschwarz | carl_baldwin, ahh, that's ok+ | 17:06 |
| *** Marga_ has joined #openstack-neutron | 17:06 | |
| jschwarz | carl_baldwin, I don't have *that* much free time so it's no biggy | 17:06 |
| carl_baldwin | Ahh, free time, I haven't that in a while. | 17:07 |
| *** aduarte has joined #openstack-neutron | 17:07 | |
| *** daneyon has quit IRC | 17:08 | |
| *** anilvenkata1 has quit IRC | 17:08 | |
| *** enikanorov2 has joined #openstack-neutron | 17:08 | |
| *** banix has joined #openstack-neutron | 17:09 | |
| *** baoli has quit IRC | 17:10 | |
| *** pritesh has joined #openstack-neutron | 17:11 | |
| *** baoli has joined #openstack-neutron | 17:12 | |
| *** claudiub has joined #openstack-neutron | 17:12 | |
| *** ivar-lazzaro has quit IRC | 17:12 | |
| *** padkrish has joined #openstack-neutron | 17:13 | |
| *** prad_ has joined #openstack-neutron | 17:14 | |
| *** tidwellr has quit IRC | 17:14 | |
| *** andyhky has quit IRC | 17:15 | |
| *** claudiub|2 has quit IRC | 17:15 | |
| *** turul has joined #openstack-neutron | 17:15 | |
| *** turul is now known as afazekas | 17:15 | |
| *** dguitarbite has joined #openstack-neutron | 17:15 | |
| *** andyhky has joined #openstack-neutron | 17:16 | |
| *** Marga_ has quit IRC | 17:16 | |
| *** prad has quit IRC | 17:17 | |
| *** tiswanso_ has quit IRC | 17:18 | |
| jaypipes | ZZelle_: sorry, had to run out :( I meant the users/left resources, yes | 17:18 |
| *** tiswanso_ has joined #openstack-neutron | 17:21 | |
| *** ollivier has joined #openstack-neutron | 17:22 | |
| *** emagana has quit IRC | 17:24 | |
| *** ihrachyshka has quit IRC | 17:24 | |
| *** krtaylor has quit IRC | 17:24 | |
| *** Marga_ has joined #openstack-neutron | 17:25 | |
| *** emagana has joined #openstack-neutron | 17:26 | |
| *** pavel_bondar has joined #openstack-neutron | 17:26 | |
| *** britthouser has joined #openstack-neutron | 17:26 | |
| *** britthou_ has joined #openstack-neutron | 17:27 | |
| openstackgerrit | Pavel Bondar proposed openstack/neutron: Refactor db_base_plugin_v2.py to use new IPAM https://review.openstack.org/153236 | 17:27 |
| openstackgerrit | John Schwarz proposed openstack/neutron: Fullstack resources should always be cleaned up https://review.openstack.org/181991 | 17:28 |
| jschwarz | carl_baldwin, ^ :) | 17:28 |
| jschwarz | amuller, ^ this also fixes one of the problems you had with fullstack a few weeks back | 17:29 |
| *** hitalia has joined #openstack-neutron | 17:29 | |
| *** baoli has quit IRC | 17:29 | |
| *** moshele has joined #openstack-neutron | 17:30 | |
| *** britthouser has quit IRC | 17:31 | |
| *** padkrish has quit IRC | 17:31 | |
| carl_baldwin | jschwarz, ack | 17:31 |
| *** nyechiel has quit IRC | 17:31 | |
| jschwarz | and I'm off | 17:32 |
| jschwarz | see you around guys :) | 17:32 |
| *** baoli has joined #openstack-neutron | 17:34 | |
| *** shashankhegde has joined #openstack-neutron | 17:34 | |
| *** krtaylor has joined #openstack-neutron | 17:36 | |
| *** yfried has quit IRC | 17:37 | |
| amuller | jschwarz: thank you | 17:37 |
| amuller | jschwarz: I'll play around with it | 17:37 |
| *** jschwarz has quit IRC | 17:37 | |
| *** rushil has quit IRC | 17:38 | |
| *** evgenyf has joined #openstack-neutron | 17:38 | |
| -openstackstatus- NOTICE: We have discovered post-upgrade issues with Gerrit affecting nova (and potentially other projects). Some changes will not appear and some actions, such as queries, may return an error. We are continuing to investigate. | 17:39 | |
| *** ChanServ changes topic to "We have discovered post-upgrade issues with Gerrit affecting nova (and potentially other projects). Some changes will not appear and some actions, such as queries, may return an error. We are continuing to investigate." | 17:39 | |
| *** moshele has quit IRC | 17:40 | |
| *** mestery has joined #openstack-neutron | 17:41 | |
| *** rushil has joined #openstack-neutron | 17:41 | |
| *** jorgem has joined #openstack-neutron | 17:41 | |
| *** [2]evgenyf has joined #openstack-neutron | 17:42 | |
| *** evgenyf has quit IRC | 17:42 | |
| *** [2]evgenyf is now known as evgenyf | 17:42 | |
| *** mturpin has joined #openstack-neutron | 17:42 | |
| *** annegentle has quit IRC | 17:43 | |
| *** baoli has quit IRC | 17:43 | |
| *** tidwellr has joined #openstack-neutron | 17:44 | |
| *** mestery has quit IRC | 17:44 | |
| *** moshele has joined #openstack-neutron | 17:45 | |
| *** rushil has quit IRC | 17:45 | |
| *** mestery has joined #openstack-neutron | 17:46 | |
| *** evgenyf has quit IRC | 17:47 | |
| *** rushil has joined #openstack-neutron | 17:47 | |
| pavel_bondar | carl_baldwin: Hi! I believe I have addressed almost all your comments. Could you please provide new portion of them? #link https://review.openstack.org//#/c/153236/ | 17:49 |
| carl_baldwin | pavel_bondar, thanks. I will soon. | 17:50 |
| carl_baldwin | pavel_bondar, sorry my last review was not comprehensive. I normally try to finish a review. | 17:51 |
| *** jorgem has quit IRC | 17:51 | |
| *** sridhar_ram has joined #openstack-neutron | 17:52 | |
| *** gridinv_ has joined #openstack-neutron | 17:53 | |
| *** moha_hunt has joined #openstack-neutron | 17:53 | |
| *** achanda has joined #openstack-neutron | 17:54 | |
| *** ivar-lazzaro has joined #openstack-neutron | 17:56 | |
| *** mestery has quit IRC | 17:56 | |
| *** prad_ is now known as prad | 17:57 | |
| *** gridinv_ has quit IRC | 17:57 | |
| *** baoli has joined #openstack-neutron | 17:58 | |
| *** dguitarbite has quit IRC | 18:00 | |
| pavel_bondar | carl_baldwin: It is +2148, -917 lines of changes, so it is almost impossible to finish review :) | 18:00 |
| carl_baldwin | pavel_bondar, good point. :) | 18:01 |
| kevinbenton | marun, amuller: https://review.openstack.org//#/c/181674/ | 18:01 |
| *** Networkn3rd has quit IRC | 18:01 | |
| *** yeungp has quit IRC | 18:02 | |
| *** ajmiller has joined #openstack-neutron | 18:02 | |
| kevinbenton | The dhcp issue that was discussed before that was incorrectly blamed for gate discomfort on Friday | 18:03 |
| amuller | kevinbenton: shame you didnt coordinate with me | 18:03 |
| amuller | kevinbenton: I'll review it this week | 18:03 |
| kevinbenton | This is something I've been sitting on for a while because I'm having a hard time coming up with something that isn't special cased | 18:04 |
| *** redbeard has quit IRC | 18:04 | |
| kevinbenton | Which is why I just started with that original patch to change when dhcp agents are scheduled | 18:05 |
| *** ijw has joined #openstack-neutron | 18:05 | |
| kevinbenton | amuller: do you have another approach? | 18:05 |
| *** ijw has quit IRC | 18:05 | |
| *** ijw has joined #openstack-neutron | 18:05 | |
| amuller | kevinbenton: I was doing design work for a dhcping based approach from the OVS agent, it would require some changes to the DHCP agent and some structural changes, and I've always liked the DHCP agent based approach better | 18:06 |
| amuller | kevinbenton: so it'd work with all plugins/agents that use the DHCP agent without special casing the OVS agent | 18:06 |
| kevinbenton | amuller: so your approach is to put all of the readiness logic into OVS agent | 18:07 |
| amuller | kevinbenton: yeah, again I don't like it as much | 18:07 |
| *** aranjan has quit IRC | 18:09 | |
| *** aranjan has joined #openstack-neutron | 18:10 | |
| kevinbenton | amuller: I tried to make mine generic to handle other things that have a composite status | 18:10 |
| *** marutikamat has joined #openstack-neutron | 18:10 | |
| kevinbenton | amuller: but this still feels a bit bolted on | 18:11 |
| amuller | kevinbenton: using call in the DHCP agent is going to slow it down massively | 18:11 |
| amuller | which is where I spent most of my thinking in the OVS agent, it's the same problem | 18:11 |
| amuller | kevinbenton: you don't need to block until the server finishes its DB work | 18:11 |
| kevinbenton | amuller: don't worry about that, that can be optimized | 18:11 |
| *** sthillma has joined #openstack-neutron | 18:11 | |
| amuller | kevinbenton: well, let's talk about that :) | 18:12 |
| *** sbalukoff has joined #openstack-neutron | 18:12 | |
| kevinbenton | amuller: the server side stuff is more what I'm worried about | 18:12 |
| *** Networkn3rd has joined #openstack-neutron | 18:12 | |
| *** annegentle has joined #openstack-neutron | 18:12 | |
| openstackgerrit | Pavel Bondar proposed openstack/neutron: Enable pluggable IPAM implementation by default https://review.openstack.org/181023 | 18:12 |
| kevinbenton | amuller: because it's creation and deletion of records for every port | 18:12 |
| *** jorgem has joined #openstack-neutron | 18:13 | |
| kevinbenton | amuller: I need to step away for a bit. Leave that comment and others on the patch | 18:13 |
| Sukhdev | csd: are you here? | 18:13 |
| *** padkrish has joined #openstack-neutron | 18:14 | |
| Sukhdev | marutikamat: are you here? | 18:14 |
| csd | Sukhdev yes | 18:14 |
| Sukhdev | csd, marutikamat : I have link ready for you | 18:14 |
| Sukhdev | csd: https://docs.google.com/presentation/d/1iIfkkbgL0bsShJ38yS8-pK_dwdNnxGptEP4yhkY-n8Q/edit?usp=sharing | 18:14 |
| Sukhdev | csd: see if you can see it? | 18:14 |
| csd | Sukhdev works. thanks | 18:15 |
| *** sripriya has joined #openstack-neutron | 18:15 | |
| amuller | kevinbenton: ttyl | 18:15 |
| amuller | kevinbenton: full stack testing smells appropriate here | 18:15 |
| amuller | kevinbenton: we just need to merge a few patches there | 18:15 |
| Sukhdev | csd: can you pass it on to marutikamat ? | 18:15 |
| Sukhdev | csd: does it make sense? | 18:15 |
| *** emagana has quit IRC | 18:15 | |
| csd | Sukhdev makes sense. | 18:16 |
| Sukhdev | csd: cool - let me know if you need anything else | 18:17 |
| *** emagana has joined #openstack-neutron | 18:17 | |
| csd | Sukhdev thanks. will ping you again on it later | 18:17 |
| *** achanda has quit IRC | 18:18 | |
| *** marutikamat has quit IRC | 18:19 | |
| *** TrevorV|Home has joined #openstack-neutron | 18:19 | |
| *** yfried has joined #openstack-neutron | 18:19 | |
| *** yfried has quit IRC | 18:19 | |
| *** baoli has quit IRC | 18:21 | |
| *** baoli has joined #openstack-neutron | 18:21 | |
| *** lennybe has joined #openstack-neutron | 18:23 | |
| *** lennybe has left #openstack-neutron | 18:23 | |
| *** enikanorov has joined #openstack-neutron | 18:23 | |
| *** obondarev_ has joined #openstack-neutron | 18:23 | |
| *** lennyb has joined #openstack-neutron | 18:23 | |
| *** sripriya has quit IRC | 18:25 | |
| *** andyhky has quit IRC | 18:25 | |
| *** ronis has quit IRC | 18:25 | |
| *** alaski has quit IRC | 18:25 | |
| *** dboik_ has quit IRC | 18:25 | |
| *** rpothier has quit IRC | 18:25 | |
| *** hanzhang has quit IRC | 18:25 | |
| *** Guest64787 has quit IRC | 18:25 | |
| *** smallbig has quit IRC | 18:25 | |
| *** morganfainberg has quit IRC | 18:25 | |
| *** clett has quit IRC | 18:25 | |
| *** aleksandr_null has quit IRC | 18:25 | |
| *** russellb has quit IRC | 18:25 | |
| *** akamyshnikova has quit IRC | 18:25 | |
| *** haleyb has quit IRC | 18:25 | |
| *** strictlyb has quit IRC | 18:25 | |
| *** ianw has quit IRC | 18:25 | |
| *** stanzgy_ has joined #openstack-neutron | 18:25 | |
| *** larsks|alt has joined #openstack-neutron | 18:26 | |
| *** emagana has quit IRC | 18:26 | |
| *** michatotol_ has quit IRC | 18:26 | |
| *** stanzgy__ has quit IRC | 18:26 | |
| *** larsks has quit IRC | 18:26 | |
| *** marrusl has quit IRC | 18:26 | |
| *** dmellado has quit IRC | 18:26 | |
| *** x58 has quit IRC | 18:26 | |
| *** phschwartz has quit IRC | 18:26 | |
| *** tkatarki1 has joined #openstack-neutron | 18:26 | |
| *** claudiub|2 has joined #openstack-neutron | 18:26 | |
| *** claudiub has quit IRC | 18:26 | |
| *** yfried has joined #openstack-neutron | 18:26 | |
| *** rushil has quit IRC | 18:26 | |
| *** tkatarki has quit IRC | 18:26 | |
| *** enikanorov_ has quit IRC | 18:26 | |
| *** obondarev has quit IRC | 18:26 | |
| *** tobias- has quit IRC | 18:26 | |
| *** marrusl has joined #openstack-neutron | 18:26 | |
| *** tkatarki1 has quit IRC | 18:26 | |
| *** tkatarki1 has joined #openstack-neutron | 18:26 | |
| *** x58 has joined #openstack-neutron | 18:26 | |
| *** sripriya has joined #openstack-neutron | 18:27 | |
| *** andyhky has joined #openstack-neutron | 18:27 | |
| *** ronis has joined #openstack-neutron | 18:27 | |
| *** alaski has joined #openstack-neutron | 18:27 | |
| *** dboik_ has joined #openstack-neutron | 18:27 | |
| *** rpothier has joined #openstack-neutron | 18:27 | |
| *** hanzhang has joined #openstack-neutron | 18:27 | |
| *** smallbig has joined #openstack-neutron | 18:27 | |
| *** morganfainberg has joined #openstack-neutron | 18:27 | |
| *** clett has joined #openstack-neutron | 18:27 | |
| *** aleksandr_null has joined #openstack-neutron | 18:27 | |
| *** russellb has joined #openstack-neutron | 18:27 | |
| *** akamyshnikova has joined #openstack-neutron | 18:27 | |
| *** haleyb has joined #openstack-neutron | 18:27 | |
| *** strictlyb has joined #openstack-neutron | 18:27 | |
| *** ianw has joined #openstack-neutron | 18:27 | |
| *** dmellado has joined #openstack-neutron | 18:27 | |
| openstackgerrit | Sam Betts proposed openstack/neutron: Add an internal DHCPv6 client for Prefix Delegation https://review.openstack.org/161381 | 18:27 |
| *** tobias- has joined #openstack-neutron | 18:27 | |
| *** jorgem has quit IRC | 18:27 | |
| *** emagana has joined #openstack-neutron | 18:28 | |
| *** phschwartz has joined #openstack-neutron | 18:28 | |
| *** yamamoto has joined #openstack-neutron | 18:28 | |
| *** larsks|alt is now known as larsks | 18:28 | |
| *** michatotol_ has joined #openstack-neutron | 18:28 | |
| *** annegentle has quit IRC | 18:29 | |
| *** jorgem has joined #openstack-neutron | 18:29 | |
| *** dan| has joined #openstack-neutron | 18:29 | |
| *** annegentle has joined #openstack-neutron | 18:29 | |
| *** tkatarki1 has quit IRC | 18:31 | |
| openstackgerrit | Ila Palanisamy proposed openstack/neutron: manual add/remove router for dvr_snat agent https://review.openstack.org/122024 | 18:31 |
| *** tkatarki has joined #openstack-neutron | 18:31 | |
| *** atuvenie has quit IRC | 18:32 | |
| *** yamamoto has quit IRC | 18:33 | |
| openstackgerrit | Ila Palanisamy proposed openstack/neutron: manual add/remove router for dvr_snat agent https://review.openstack.org/122024 | 18:33 |
| *** Sukhdev has quit IRC | 18:35 | |
| *** ZZelle has joined #openstack-neutron | 18:36 | |
| *** zhenguo has quit IRC | 18:37 | |
| *** ollivier has quit IRC | 18:38 | |
| *** annegentle has quit IRC | 18:40 | |
| *** yamahata has quit IRC | 18:40 | |
| *** nlahouti has quit IRC | 18:41 | |
| *** annegentle has joined #openstack-neutron | 18:41 | |
| *** padkrish has quit IRC | 18:41 | |
| *** padkrish has joined #openstack-neutron | 18:42 | |
| openstackgerrit | Ryan Tidwell proposed openstack/neutron: Block subnet create when a network hosts subnets allocated from different pools https://review.openstack.org/180299 | 18:42 |
| *** ashishjain has joined #openstack-neutron | 18:46 | |
| *** nlahouti has joined #openstack-neutron | 18:46 | |
| *** padkrish has quit IRC | 18:46 | |
| ashishjain | Hello | 18:46 |
| *** padkrish has joined #openstack-neutron | 18:46 | |
| ashishjain | I am stuck with a weird problem which seems to be related to payload size | 18:47 |
| ashishjain | I am using neutron networking. | 18:47 |
| ashishjain | Is there anyone who can provide some tips to debug this problem out. | 18:47 |
| openstackgerrit | Carl Baldwin proposed openstack/neutron-specs: Implement routing networks in Neutron https://review.openstack.org/172244 | 18:48 |
| ashishjain | I keep getting 504 error whenever the payload size is big | 18:49 |
| *** dan| has quit IRC | 18:49 | |
| ashishjain | whereas I get the response when the payload size is bare miniumu | 18:50 |
| ashishjain | *minimum | 18:50 |
| openstackgerrit | Carl Baldwin proposed openstack/neutron-specs: Introduce address scopes https://review.openstack.org/180267 | 18:51 |
| *** baoli has quit IRC | 18:51 | |
| *** harlowja has quit IRC | 18:52 | |
| *** harlowja has joined #openstack-neutron | 18:52 | |
| *** ronis has quit IRC | 18:54 | |
| *** dan_ has joined #openstack-neutron | 18:54 | |
| *** dan_ is now known as Guest25580 | 18:55 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: Move full-stack logs post-tests https://review.openstack.org/180882 | 18:58 |
| *** ijw has quit IRC | 18:58 | |
| *** ijw has joined #openstack-neutron | 18:59 | |
| openstackgerrit | Carl Baldwin proposed openstack/neutron: Make it clear the rfe tag is lower-case https://review.openstack.org/182017 | 19:00 |
| *** britthou_ has quit IRC | 19:01 | |
| *** nyechiel has joined #openstack-neutron | 19:02 | |
| *** jschwarz has joined #openstack-neutron | 19:03 | |
| *** ijw has quit IRC | 19:03 | |
| *** crose has joined #openstack-neutron | 19:03 | |
| *** yamahata has joined #openstack-neutron | 19:03 | |
| jschwarz | carl_baldwin, re: your comment on https://review.openstack.org/#/c/180882/4..5/neutron/tests/contrib/post_test_hook.sh | 19:04 |
| *** emagana has quit IRC | 19:05 | |
| *** emagana has joined #openstack-neutron | 19:05 | |
| *** britthouser has joined #openstack-neutron | 19:06 | |
| *** Marga_ has quit IRC | 19:11 | |
| carl_baldwin | jschwarz, did you want to discuss it or are you just pointing out your reply? | 19:12 |
| *** dguerri is now known as _dguerri | 19:12 | |
| jschwarz | carl_baldwin, pointing it out, seeing if you want to discuss | 19:16 |
| *** nyechiel has quit IRC | 19:16 | |
| carl_baldwin | jschwarz, I think we should get some other opinions. | 19:17 |
| jschwarz | okies :) | 19:17 |
| *** dave-mccowan has quit IRC | 19:17 | |
| *** smccully has quit IRC | 19:18 | |
| *** yeungp has joined #openstack-neutron | 19:20 | |
| *** jschwarz has quit IRC | 19:21 | |
| kevinbenton | carl_baldwin: bugs for features? I thought we hadn't decided to can specs yet | 19:25 |
| carl_baldwin | kevinbenton, where've you been all week-end? ;) https://review.openstack.org/#/c/177342/ | 19:27 |
| *** erkules has quit IRC | 19:27 | |
| carl_baldwin | tl;dr specs are still in, no "deadline". rfe bugs are also in. | 19:27 |
| *** jprovazn has quit IRC | 19:29 | |
| *** subscope has joined #openstack-neutron | 19:29 | |
| *** dave-mccowan has joined #openstack-neutron | 19:31 | |
| *** mchestr has quit IRC | 19:32 | |
| *** mchestr has joined #openstack-neutron | 19:32 | |
| kevinbenton | What is an rfe? This acronym suddenly showed up everywhere in the last two weeks. | 19:37 |
| kevinbenton | carl_baldwin: ^^ | 19:37 |
| amuller | request for feature | 19:37 |
| amuller | kevinbenton: ^ | 19:37 |
| amuller | request feature enhancement | 19:38 |
| carl_baldwin | kevinbenton, Honestly, I don't know. Request for enhancement? | 19:38 |
| *** _dguerri is now known as dguerri | 19:38 | |
| kevinbenton | carl_baldwin: so then a spec would be filed to address an rfe? | 19:38 |
| carl_baldwin | kevinbenton, that sounds about right. I think that small enhancements may not require a full blown spec. Requirements for specs seem to have loosened a bit where some detail may not be required if not requested by reviewers. | 19:39 |
| kevinbenton | carl_baldwin: makes sense. I knew specs were getting a huge diet | 19:40 |
| kevinbenton | carl_baldwin: I just didn't know about these rfes | 19:40 |
| *** hareeshp has quit IRC | 19:40 | |
| kevinbenton | Henceforth pronounced 'refees' | 19:40 |
| carl_baldwin | :) | 19:41 |
| *** erkules has joined #openstack-neutron | 19:42 | |
| *** Marga_ has joined #openstack-neutron | 19:42 | |
| *** bjornar__ has joined #openstack-neutron | 19:44 | |
| *** hitalia has quit IRC | 19:45 | |
| *** hareeshp has joined #openstack-neutron | 19:45 | |
| *** Marga_ has quit IRC | 19:46 | |
| *** britthouser has quit IRC | 19:47 | |
| *** aranjan has quit IRC | 19:48 | |
| *** matrohon has joined #openstack-neutron | 19:50 | |
| *** rushil has joined #openstack-neutron | 19:50 | |
| *** aranjan_ has joined #openstack-neutron | 19:50 | |
| *** aranjan_ has quit IRC | 19:53 | |
| *** bjornar__ has quit IRC | 19:58 | |
| *** lennyb has left #openstack-neutron | 20:00 | |
| *** hareeshp has quit IRC | 20:01 | |
| *** dims has joined #openstack-neutron | 20:03 | |
| *** rushil has quit IRC | 20:05 | |
| *** tiswanso_ has quit IRC | 20:06 | |
| *** mwagner_lap has quit IRC | 20:07 | |
| *** sdake_ has joined #openstack-neutron | 20:10 | |
| *** nihilifer has quit IRC | 20:11 | |
| *** doude has quit IRC | 20:11 | |
| *** doude has joined #openstack-neutron | 20:11 | |
| *** daneyon_ has quit IRC | 20:12 | |
| *** daneyon has joined #openstack-neutron | 20:13 | |
| *** sdake has quit IRC | 20:13 | |
| *** hitalia has joined #openstack-neutron | 20:14 | |
| *** nlahouti has quit IRC | 20:14 | |
| enikanorov2 | gerrit seems to be having problems | 20:15 |
| salv-orlando | enikanorov2: it's telling you that it's time to sleep | 20:16 |
| *** b3nt_pin has joined #openstack-neutron | 20:17 | |
| enikanorov2 | salv-orlando: i'm in pdt | 20:19 |
| enikanorov2 | but i'd love to have some after-lunch sleep | 20:19 |
| enikanorov2 | like sleep(7200) | 20:20 |
| ZZelle | jaypipes, hi | 20:21 |
| openstackgerrit | Cedric Brandily proposed openstack/python-neutronclient: Allow setting router's external ip(s) https://review.openstack.org/181682 | 20:23 |
| jaypipes | ZZelle: hey | 20:24 |
| *** aranjan has joined #openstack-neutron | 20:24 | |
| *** b3nt_pin has quit IRC | 20:25 | |
| *** b3nt_pin has joined #openstack-neutron | 20:26 | |
| *** Marga_ has joined #openstack-neutron | 20:29 | |
| *** padkrish has quit IRC | 20:34 | |
| ZZelle | jaypipes, did you get your response on quotas? | 20:35 |
| jaypipes | ZZelle: no, I don't think so. I don't understand where in Neutron the reserve/allocate/commit-or-rollback operations are performed | 20:38 |
| openstackgerrit | Carl Baldwin proposed openstack/neutron-specs: Introduce address scopes https://review.openstack.org/180267 | 20:38 |
| ZZelle | jaypipes, only quota limits are persisted | 20:38 |
| jaypipes | ZZelle: I don't understand what that means. Could you elaborate? | 20:39 |
| *** padkrish has joined #openstack-neutron | 20:40 | |
| *** baoli has joined #openstack-neutron | 20:40 | |
| salv-orlando | jaypipes: are you talking about quotas? | 20:40 |
| ZZelle | jaypipes, of course, i am trying to remember where it's done in neutron | 20:41 |
| openstackgerrit | Carl Baldwin proposed openstack/neutron-specs: Introduce address scopes https://review.openstack.org/180267 | 20:41 |
| salv-orlando | jaypipes: I believe there is only one neutron instance in the world that has the concept of reservations. The one in my local devstack setup. | 20:42 |
| *** rushil has joined #openstack-neutron | 20:42 | |
| *** baoli has quit IRC | 20:44 | |
| *** openstackgerrit_ has joined #openstack-neutron | 20:44 | |
| *** TrevorV|Home has quit IRC | 20:44 | |
| jaypipes | salv-orlando: so neutron has no RESERVE .. do some stuff and either COMMIT or ROLLBACK the reserved amount? | 20:44 |
| *** baoli has joined #openstack-neutron | 20:44 | |
| *** cbouch has joined #openstack-neutron | 20:45 | |
| salv-orlando | jaypipes: nope. Indeed it is perfectly possible to gain limit x N allocated quota if you are lucky and there are N api servers | 20:45 |
| ZZelle | jaypipes, nothing is persisted except limits ... when neutron checks a limit, neutron performs a db query | 20:45 |
| ZZelle | jaypipes, counting allocated resources is done by: https://github.com/openstack/neutron/blob/master/neutron/quota.py#L316-L329 | 20:46 |
| jaypipes | hmmm.... no wonder there's concurrency/deadlock issues... | 20:46 |
| ZZelle | jaypipes, for sure, there is | 20:47 |
| jaypipes | ok, thx guys! that answers my questions! | 20:47 |
| jaypipes | ZZelle: it was in regards to https://review.openstack.org/#/c/179955/4/neutron/ipam/subnet_alloc.py | 20:47 |
| *** TrevorV|home has joined #openstack-neutron | 20:47 | |
| ZZelle | jaypipes, that's why i am not sure we should bother with lockmode when checking quotas in subnet_alloc | 20:47 |
| jaypipes | ZZelle: I'm pretty sure the changes there won't resolve any lock contention and concurrency issues due to the quota system being racy | 20:48 |
| *** achanda has joined #openstack-neutron | 20:48 | |
| *** merooney has quit IRC | 20:48 | |
| *** padkrish has quit IRC | 20:48 | |
| salv-orlando | jaypipes, ZZelle | 20:49 |
| salv-orlando | sorry I should write something... | 20:49 |
| *** padkrish has joined #openstack-neutron | 20:49 | |
| ZZelle | jaypipes, for sure, we need something to lock the db or version quota if we want to "strongly" enforce quotas | 20:49 |
| jaypipes | ZZelle: what I don't understand is by returning an IpamSubnet() object, what does that do exactly? When does the database (or other datastore) get wrtitten/updated? | 20:49 |
| *** padkrish has quit IRC | 20:49 | |
| salv-orlando | for allocating subnet from pools quotas are kind of 'unique'. They're not managed by the quota engine. | 20:49 |
| *** padkrish has joined #openstack-neutron | 20:50 | |
| ZZelle | salv-orlando, could you elaborate? | 20:50 |
| jaypipes | salv-orlando: I don't see anywhere in https://review.openstack.org/#/c/179955/4/neutron/ipam/subnet_alloc.py where concurrent updates are handled... | 20:50 |
| salv-orlando | not that I second that, but the team working on that strongly felt that quota mechanisms did not satisfy their main requirement, which was to allow limits on a subnet pool basis | 20:50 |
| salv-orlando | basically with things like networks, or ports, you have per-tenant limits. | 20:51 |
| *** merooney has joined #openstack-neutron | 20:51 | |
| salv-orlando | In the case of subnet pool, the limit belongs to the subnet pool itsel.f | 20:51 |
| jaypipes | salv-orlando: for instance, in lines 109-110 introduce a race where multiple writers can get the same subnet from the pool. | 20:51 |
| jaypipes | salv-orlando: so where is the "update" to the backend data store exactly? | 20:51 |
| *** Sukhdev has joined #openstack-neutron | 20:52 | |
| jaypipes | salv-orlando: this code returns an IpamSubnet object, but I don't see where that object is saved to some persistent store? | 20:52 |
| salv-orlando | jaypipes: just a sec, switching to git.openstack.org, gerrit seems to not work or perhaps my cache needs to refreshed | 20:52 |
| jaypipes | k | 20:52 |
| ZZelle | jaypipes, salv-orlando, are you able to open the gerrit link ... i am not :( | 20:52 |
| *** claudiub has joined #openstack-neutron | 20:53 | |
| jaypipes | ZZelle: I've had it open for a while... don't want to refresh :) | 20:53 |
| ZZelle | jaypipes, so i follow salv proposal ! | 20:53 |
| carl_baldwin | gerrit seems to have lost its object store. | 20:53 |
| jaypipes | salv-orlando, ZZelle: to make things simple, I am referring to this part of _allocat_any_subnet(): | 20:53 |
| jaypipes | if not gateway_ip: | 20:53 |
| jaypipes | gateway_ip = subnet.network + 1 | 20:53 |
| jaypipes | return IpamSubnet(request.tenant_id, | 20:53 |
| jaypipes | request.subnet_id, | 20:53 |
| jaypipes | subnet.cidr, | 20:53 |
| jaypipes | gateway_ip=gateway_ip, | 20:53 |
| jaypipes | allocation_pools=None) | 20:53 |
| jaypipes | salv-orlando, ZZelle: that block is within a session.begin(), but there's no update/insert anywhere. | 20:54 |
| ZZelle | jaypipes, the lock is done on subnetpool table | 20:54 |
| jaypipes | ZZelle: where? | 20:54 |
| salv-orlando | that is a transfer object, the write is still driven by db_base_plugin_v2 | 20:54 |
| salv-orlando | jaypipes: navigating the code... just a sec | 20:54 |
| jaypipes | yeck | 20:54 |
| *** amuller has quit IRC | 20:55 | |
| ZZelle | jaypipes, in _lock_subnetpool which ensures at most one cidr is allocated per subnetpool (assuming lockmode is working) | 20:55 |
| salv-orlando | http://git.openstack.org/cgit/openstack/neutron/tree/neutron/db/db_base_plugin_v2.py#n1298 | 20:55 |
| *** claudiub|2 has quit IRC | 20:55 | |
| salv-orlando | jaypipes: as I told several times during design and implementation, we had to balance the need of having this feature with the strong java-ish orientation of involved developers. | 20:56 |
| salv-orlando | kindof wherever you can do an if, use plymorphism ;) | 20:56 |
| jaypipes | lol | 20:56 |
| salv-orlando | anyway, at least we have proper objects and not random dicts passed around | 20:56 |
| jaypipes | true | 20:56 |
| salv-orlando | I tend to look at the bright side, always | 20:57 |
| salv-orlando | otherwise I would have already committed suicide | 20:57 |
| jaypipes | salv-orlando: k, so http://git.openstack.org/cgit/openstack/neutron/tree/neutron/db/db_base_plugin_v2.py#n1176 | 20:57 |
| jaypipes | that whole method has a crap-ton of race conditions in it that won't be solved with a FOR UPDATE clause. | 20:57 |
| salv-orlando | yes, that is the method that specifically stores the subnet on the database | 20:57 |
| jaypipes | salv-orlando: it needs a @retry_on_deadlock on it. | 20:58 |
| ZZelle | jaypipes, i didn't see that's one the lock should be done on a parent transaction | 20:58 |
| jaypipes | also, all of the validation calls should be done *outside* of the DB transaction. | 20:58 |
| *** Marga_ has quit IRC | 20:58 | |
| salv-orlando | jaypipes: should it be applied there only or to the method where the wrapping transaction begins? | 20:58 |
| *** Marga_ has joined #openstack-neutron | 20:59 | |
| jaypipes | ZZelle: SELECT FOR UPDATE will not work like you think on Galera :) thus the need for a retry_on_deadlock decorator. | 20:59 |
| jaypipes | salv-orlando: it should be applied around an atomic unit of work. | 20:59 |
| salv-orlando | jaypipes: ZZelle knows that. he's implemented your non locking thin on vni allocations | 20:59 |
| ZZelle | jaypipes, for the moment, it's not working at all | 20:59 |
| salv-orlando | (opening another can of worms) | 20:59 |
| jaypipes | salv-orlando: which, unless I'm mistaken, is just *part* of _save_subnet() | 21:00 |
| *** setuid has quit IRC | 21:00 | |
| enikanorov2 | jaypipes: around outer-most transaction, indeed | 21:00 |
| *** tflynn has quit IRC | 21:00 | |
| ZZelle | enikanorov2, which seems to imply a lot of rework :'( | 21:00 |
| ZZelle | for subnetpool | 21:00 |
| jaypipes | enikanorov: you could, yes, but I'd advise removing say, lines 1183-1193 from the DB transaction and putting that validation code outside of the DB transaction entirely. | 21:00 |
| salv-orlando | enikanorov2, jaypipes: yeah because with active/active clusters obviously write-set certification occurs at commit time | 21:00 |
| enikanorov2 | jaypipes: agree on that | 21:01 |
| *** jbell8 has quit IRC | 21:01 | |
| salv-orlando | yeah I thinik I told that to the author during code review as well.... | 21:01 |
| jaypipes | ZZelle: no, actually shouldn't be too much work. just create a little internal function inside of _save_subnet() and wrap that internal func with @retry_on_deadlock | 21:01 |
| jaypipes | ZZelle: and then you actually wouldn't need any changes at all in the IPAM module. | 21:02 |
| *** mwagner_lap has joined #openstack-neutron | 21:02 | |
| jaypipes | ZZelle: basically take lines 1195-1222 and put that into an internal func, wrapped by the retry_on_deadlokc | 21:02 |
| jaypipes | ZZelle: and forget about locking :) cuz it don't work like you think it does. | 21:02 |
| jaypipes | at least, in galera-land. for better or worse. | 21:03 |
| salv-orlando | ZZelle: there's only one more twist about subnet pools you need to considers. | 21:03 |
| salv-orlando | consd | 21:03 |
| salv-orlando | consider. crap. | 21:03 |
| *** aranjan has quit IRC | 21:03 | |
| *** setuid has joined #openstack-neutron | 21:03 | |
| salv-orlando | The SubnetPool has a relationship with SubnetPoolPrefixes which are all the allocated CIDRs on that pool | 21:03 |
| salv-orlando | but SubnetPoolPrefixes has no relationship whatsoever with Subnets | 21:04 |
| *** jbell8 has joined #openstack-neutron | 21:04 | |
| *** sdake has joined #openstack-neutron | 21:04 | |
| salv-orlando | this means that by enclosing the transaction in _save_subnet we need to be careful about ensuring consistency in case of rollbacks. | 21:04 |
| salv-orlando | This is I believe, what is called data de-normalization | 21:05 |
| salv-orlando | when I was in college if I did something like that I would fail the exam | 21:05 |
| salv-orlando | but that was early 21th century, we still used things like IBM DB2 | 21:05 |
| salv-orlando | I guess things are changed now and denormalization is ok | 21:05 |
| *** hichihara has joined #openstack-neutron | 21:05 | |
| *** ashishjain has quit IRC | 21:06 | |
| ZZelle | jaypipes, why do we need to perform check outside of enclosing transaction? | 21:06 |
| jaypipes | salv-orlando: denormalization is a separate concept :) | 21:06 |
| ZZelle | jaypipes, changing isolation level is not enough? | 21:06 |
| jaypipes | ZZelle: just a best practice. you want to reduce the number of operations performed while in a BEGIN/COMMIT transactional block, and none of those validation operations need the database. | 21:07 |
| *** botang has quit IRC | 21:07 | |
| jaypipes | ZZelle: the isolation level thing is well, a long conversation. just ask enikanorov ;) | 21:07 |
| * carl_baldwin hopes all the good feedback is on the review when gerrit is happy again. | 21:07 | |
| ZZelle | jaypipes, i know, i (try to) follow it | 21:07 |
| jaypipes | carl_baldwin: so it seems that gerrit is particular about what it doesn't like... | 21:07 |
| *** jbell8 has quit IRC | 21:08 | |
| jaypipes | carl_baldwin: other patches are working fine for me, but I'm getting a jgit object reference failed for ZZelle's patch in question :( | 21:08 |
| *** sdake_ has quit IRC | 21:08 | |
| *** botang has joined #openstack-neutron | 21:08 | |
| carl_baldwin | jaypipes, I'm getting more and more unhappy gerrit as time goes on. :( | 21:08 |
| *** wwriverrat has joined #openstack-neutron | 21:08 | |
| *** aranjan has joined #openstack-neutron | 21:08 | |
| *** rushil has quit IRC | 21:09 | |
| HenryG | My gerrit problems went away after I logged out of gerrit and closed the browser. | 21:09 |
| ZZelle | infra team is trying to find what is not working in new gerrit version | 21:09 |
| anteaya | and discussing a rollback | 21:09 |
| *** Marga_ has quit IRC | 21:10 | |
| *** moshele has quit IRC | 21:10 | |
| ZZelle | anteaya, outch | 21:10 |
| anteaya | yes | 21:10 |
| anteaya | but it is on the table | 21:10 |
| salv-orlando | jaypipes: that was a half-rant. I was merely saying that by changing the transaction scope you might end up having two distinct transactions and if the 2nd fails manual actions should be taken to undo the 1st | 21:10 |
| *** Marga_ has joined #openstack-neutron | 21:10 | |
| *** annegent_ has joined #openstack-neutron | 21:10 | |
| jaypipes | salv-orlando: ah, sure, though I'd of course using a single trx to begin with ;) | 21:11 |
| matrohon | marun : hi | 21:12 |
| jaypipes | anteaya: you like my suggestion for a new kernel tag in https://review.openstack.org/#/c/180112/2/reference/tags/kernel_compute.rst ? :) | 21:12 |
| marun | matrohon: hi | 21:12 |
| *** hareeshp has joined #openstack-neutron | 21:12 | |
| * anteaya looks | 21:13 | |
| *** apuimedo has joined #openstack-neutron | 21:13 | |
| * jaypipes wonders if the Orville Redenbacher popcorn joke will translate...\ | 21:13 | |
| matrohon | marun : I was using UT in vpnaas and I found an issue : https://bugs.launchpad.net/neutron/+bug/1453965 | 21:13 |
| openstack | Launchpad bug 1453965 in neutron "extension UT not run anymore" [Undecided,New] | 21:13 |
| anteaya | it does for me | 21:13 |
| *** tkatarki has quit IRC | 21:13 | |
| anteaya | jaypipes: when you make a tag suggestion with pug in it be sure to alert me to that one as well | 21:14 |
| matrohon | marun : I discussed it with pcm_, he wasn't aware of the issue | 21:14 |
| jaypipes | anteaya: haha :) | 21:14 |
| anteaya | :) | 21:14 |
| anteaya | extra points if you get them in a stroller | 21:14 |
| *** annegentle has quit IRC | 21:15 | |
| *** baoli has quit IRC | 21:15 | |
| *** baoli has joined #openstack-neutron | 21:15 | |
| *** hareeshp has quit IRC | 21:17 | |
| *** marrusl has quit IRC | 21:17 | |
| *** s3wong has joined #openstack-neutron | 21:17 | |
| *** Marga_ has quit IRC | 21:17 | |
| *** echirivella has joined #openstack-neutron | 21:18 | |
| *** Marga_ has joined #openstack-neutron | 21:18 | |
| matrohon | marun : I was wondering if this change was involved : https://github.com/openstack/neutron/commit/1105782e3914f601b8f4be64939816b1afe8fb54 | 21:18 |
| * echirivella Hi, everyone | 21:19 | |
| sc68cal | marun: I added some info to your etherpad about a cross project session that sdague and I are moderating - https://libertydesignsummit.sched.org/event/85141d833ce535dd631fea4b0ca69ac3#.VVEcUhdFORs | 21:19 |
| marun | sc68cal: ugh | 21:20 |
| ZZelle | salv-orlando, carl_baldwin, how do we take into account allow_overlapping_ips = False when using subnetpools? | 21:20 |
| marun | sc68cal: one of the biggest gaps has nothing to do with neutron imho | 21:20 |
| marun | sc68cal: -> nova's networking api assuming synchronous operations | 21:21 |
| salv-orlando | ZZelle: I do not remember on top of my head, carl_baldwin surely knows. I think his thought is that, if I'm not wrong, that option is useless with subnet pools. It should be deprecated and neutron APIs default to subnet pools always | 21:21 |
| salv-orlando | but wait for an answers from carl_baldwin as I might not be precise | 21:21 |
| sc68cal | marun: understood - I think we're just collecting things - there's supposed to be another session on thrusday to allow people to digest | 21:23 |
| *** absubram has quit IRC | 21:23 | |
| ZZelle | salv-orlando, carl_baldwin, it's done here: https://github.com/openstack/neutron/blob/master/neutron/db/db_base_plugin_v2.py#L1193 ... but seems not to support race conditions | 21:23 |
| *** padkrish has quit IRC | 21:23 | |
| *** padkrish has joined #openstack-neutron | 21:24 | |
| marun | sc68cal: fair enough. | 21:24 |
| sc68cal | marun: the one on thursday is https://libertydesignsummit.sched.org/event/495f67e8acd369d1ac459b6ca9882340#.VVEeHRdFORs | 21:24 |
| anteaya | we are trying something with gerrit | 21:25 |
| marun | sc68cal: the use case discussion is actually more about trying to push for more user-focused, but the title doesn't capture that very well. | 21:25 |
| salv-orlando | ZZelle: are you looking at races in subnet allocations when overlapping ips and cidr are forbidden? | 21:25 |
| anteaya | jusg in case make sure you have updated local copies of any changes made since friday | 21:25 |
| salv-orlando | ZZelle: in that case there's a very old bug for it | 21:25 |
| anteaya | s/jusg/just | 21:25 |
| *** vhoward has left #openstack-neutron | 21:26 | |
| ZZelle | salv-orlando, i am trying to understand how implicit/explicit subnetpool allocation works | 21:26 |
| tidwellr | salv-orlando ZZelle: allow_overlapping_ips = False is really incompatible with subnetpools. We would like to deprecate it | 21:27 |
| *** padkrish_ has joined #openstack-neutron | 21:27 | |
| *** padkrish_ has joined #openstack-neutron | 21:27 | |
| ZZelle | tidwellr, according to code, it's compatible (but not thread-safe) | 21:28 |
| *** padkrish has quit IRC | 21:28 | |
| *** padkrish has joined #openstack-neutron | 21:29 | |
| ZZelle | tidwellr, imo, we should remove it in favor of a required subnetpool/address scope | 21:29 |
| tidwellr | ZZelle: it creates some strange behavior, but not explicitly (it should be imo) | 21:29 |
| tidwellr | *explicitly blocked | 21:29 |
| *** marrusl has joined #openstack-neutron | 21:29 | |
| marun | sc68cal: thanks for the info, though! | 21:29 |
| tidwellr | ZZelle: agreed, address scopes really make allow_overlapping_ips obsolete | 21:30 |
| *** matrohon has quit IRC | 21:30 | |
| *** julim has quit IRC | 21:31 | |
| tidwellr | ZZelle: see https://bugs.launchpad.net/neutron/+bug/1438320, we have a bug for allow_overlapping_ips = False + subnetpools | 21:32 |
| openstack | Launchpad bug 1438320 in neutron "Subnet pool created should be blocked when allow_overlapping_ips=False" [Medium,New] - Assigned to Ryan Tidwell (ryan-tidwell) | 21:32 |
| *** padkrish_ has quit IRC | 21:32 | |
| *** spandhe has joined #openstack-neutron | 21:32 | |
| ZZelle | tidwellr, imo, it's not a bug but the expected behavior as allow_overlapping_ips = False ensures non-overlapping cidrs even between subnet pools | 21:33 |
| carl_baldwin | salv-orlando isn't far off. I was probing to find out how prolifically allow_overlapping_ips is being used. If it is out there in the wild, I wouldn't mind migrating them to use subnet pools. | 21:34 |
| *** emagana has quit IRC | 21:34 | |
| *** crose has quit IRC | 21:34 | |
| *** emagana has joined #openstack-neutron | 21:35 | |
| *** gildub has joined #openstack-neutron | 21:35 | |
| *** LeanReox has joined #openstack-neutron | 21:36 | |
| *** rpothier has quit IRC | 21:36 | |
| tidwellr | ZZelle: You make a good point, what I was trying to achieve was some notion of address scopes via subnetpools. The best way to proceed is probably to just nix this bug and move forward with address scopes | 21:37 |
| carl_baldwin | ZZelle: +1 It is a good point. I still think the option could eventually be deprecated but I think we can give that some more thought. | 21:39 |
| *** TrevorV|home has quit IRC | 21:39 | |
| ZZelle | tidwellr, when default_ipv4_subnet_pool != None, we can allocate subnet only in default_ipv4_subnet_pool? or also outside with an explicit cidr? | 21:39 |
| tidwellr | ZZelle: when default_ipv4_subnet_pool != None, creating a subnet without an explicit subnetpool_id causes it to default to the supplied subnetpool. You can still use other subnetpools | 21:41 |
| *** achanda has quit IRC | 21:43 | |
| *** baoli has quit IRC | 21:44 | |
| tidwellr | ZZelle: I always speak about not supplying a subnetpool_id as using the "null" or "implicit" subnetpool. default_ipv4_subnet_pool just overrides the "implcit" subnetpool | 21:44 |
| ZZelle | tidwellr, ok | 21:44 |
| *** erkules has quit IRC | 21:45 | |
| *** erkules has joined #openstack-neutron | 21:45 | |
| carl_baldwin | ZZelle: When the default is set, you *must* send None or another pool id to the API request in order to not use it. I updated the python-neutronclient to allow doing this. | 21:46 |
| *** rushil has joined #openstack-neutron | 21:48 | |
| *** rushil has quit IRC | 21:48 | |
| LeanReox | guys, sorry does anybody knows from where openvswitch gets the TAG number for a port ? | 21:48 |
| ZZelle | carl_baldwin, it's not fully obvious as neutron behavior on subnet create without subnetpool-id depends on neutron default_ipX4_subnet_pool options | 21:48 |
| LeanReox | cause ive searched accross the whole neutron database, and its seems not to be anywhere | 21:49 |
| LeanReox | and its not the same as the segmenteation id from the net | 21:49 |
| *** padkrish has quit IRC | 21:49 | |
| *** padkrish has joined #openstack-neutron | 21:50 | |
| *** sdake has quit IRC | 21:50 | |
| *** padkrish_ has joined #openstack-neutron | 21:51 | |
| *** baoli has joined #openstack-neutron | 21:52 | |
| *** padkrish has quit IRC | 21:52 | |
| kevinbenton | LeanReox: it's locally generated by the OVS agent | 21:52 |
| *** annegent_ has quit IRC | 21:52 | |
| kevinbenton | LeanReox: it's only used to isolate ports from different networks | 21:52 |
| kevinbenton | LeanReox: as you observed it has no relation to the segmentation id | 21:52 |
| *** sdake has joined #openstack-neutron | 21:53 | |
| *** annegentle has joined #openstack-neutron | 21:53 | |
| kevinbenton | LeanReox: and the tag for one network may be different between compute nodes | 21:53 |
| LeanReox | kevinbenton, perfect, do you know what function uses to generate this tag | 21:53 |
| kevinbenton | LeanReox: one sec | 21:53 |
| LeanReox | kevinbenton, cause its the same on every port on the same subnet | 21:53 |
| kevinbenton | LeanReox: right | 21:53 |
| kevinbenton | LeanReox: https://github.com/openstack/neutron/blob/master/neutron/plugins/openvswitch/agent/ovs_neutron_agent.py#L518 | 21:54 |
| carl_baldwin | ZZelle: Do you suggest an alternative? | 21:54 |
| kevinbenton | LeanReox: i need to step away, feel free to ping me with messages and i can respond later | 21:55 |
| *** pece has quit IRC | 21:55 | |
| ZZelle | carl_baldwin, nope, but we should publish it in salv-orlando /features API :) | 21:55 |
| *** merooney has quit IRC | 21:56 | |
| carl_baldwin | ZZelle: Fair enough. | 21:56 |
| LeanReox | ok kevinbenton great, il leave you another one, if my database is unchanged and i restart the agents, why will be possibly the reason for the agent not to tag an interface and leave it on the dead vlan | 21:57 |
| LeanReox | kevinbenton, the 4095 tag | 21:57 |
| *** emagana has quit IRC | 22:00 | |
| *** botang has quit IRC | 22:00 | |
| *** annegentle has quit IRC | 22:00 | |
| *** shwetaap has quit IRC | 22:01 | |
| *** Marga_ has quit IRC | 22:04 | |
| *** Marga_ has joined #openstack-neutron | 22:04 | |
| *** baoli has quit IRC | 22:05 | |
| *** emagana has joined #openstack-neutron | 22:05 | |
| *** kbringard has quit IRC | 22:06 | |
| *** hitalia has quit IRC | 22:06 | |
| *** subscope has quit IRC | 22:09 | |
| *** aranjan has quit IRC | 22:09 | |
| *** banix has quit IRC | 22:12 | |
| *** moshele has joined #openstack-neutron | 22:13 | |
| *** jorgem has quit IRC | 22:16 | |
| *** cbouch has quit IRC | 22:18 | |
| *** subscope has joined #openstack-neutron | 22:18 | |
| *** cbouch has joined #openstack-neutron | 22:18 | |
| *** rkukura_ has joined #openstack-neutron | 22:19 | |
| *** Networkn3rd has quit IRC | 22:20 | |
| *** claudiub has quit IRC | 22:20 | |
| *** rkukura has quit IRC | 22:22 | |
| *** rkukura_ is now known as rkukura | 22:22 | |
| *** pcm_ has quit IRC | 22:23 | |
| *** jgrimm is now known as zz_jgrimm | 22:24 | |
| *** rwsu has quit IRC | 22:25 | |
| *** padkrish_ has quit IRC | 22:26 | |
| *** padkrish has joined #openstack-neutron | 22:27 | |
| *** padkrish has quit IRC | 22:27 | |
| *** padkrish has joined #openstack-neutron | 22:28 | |
| *** Networkn3rd has joined #openstack-neutron | 22:28 | |
| *** sripriya has quit IRC | 22:33 | |
| enikanorov2 | LeanReox: it could be that the port has admin_state_up = False or the port doesn't exist in the DB | 22:34 |
| *** sripriya has joined #openstack-neutron | 22:34 | |
| openstackgerrit | Carl Baldwin proposed openstack/neutron-specs: Introduce address scopes https://review.openstack.org/180267 | 22:35 |
| *** marrusl has quit IRC | 22:35 | |
| LeanReox | enikanorov2, in the ovs db -> conf.db ? or the neutron db ? | 22:36 |
| enikanorov2 | LeanReox: in neutron db. Another reason for it could be that the port is not bound | 22:36 |
| *** zhhuabj_ has quit IRC | 22:37 | |
| LeanReox | enikanorov2, ok ! and the admin_state_up = False, can be due to what ? what puts a a port in admin_state_up = False ? | 22:37 |
| enikanorov2 | LeanReox: you need to enable debug logging, there should be a reason why particular port is put to dead vlan | 22:37 |
| enikanorov2 | LeanReox: it only could be due to user's action | 22:37 |
| LeanReox | enikanorov2, can be a report_interval or dead lbaas agent, causing this ? or a dead neutron-plugin-ovs-agent ? | 22:37 |
| enikanorov2 | LeanReox: yes, dead ovs agent could be a reason of not bound port | 22:38 |
| LeanReox | enikanorov2, dead may be proccess not running or the agent down on the database = admin_state_up = 0 , or neutron agent-list XXX | 22:39 |
| *** zhhuabj_ has joined #openstack-neutron | 22:40 | |
| enikanorov2 | dead means that neutron-server doesn't receive heartbeats from ovs agent, that could be due to agent process is dead or message queue communication issues (failing rabbitmq) | 22:40 |
| LeanReox | enikanorov2, great, thats what i though, so can a agent under heavy load, cant keep up and report to the server and be marked down, and then up causing lots of ports in the dead vlan? (im tallking always about iceouse release) | 22:42 |
| enikanorov2 | LeanReox: it dipends. What kind of port do you see getting dead vlan? Is it regular VM port or L3/DHCP port? | 22:43 |
| LeanReox | most of them lbaas ports enikanorov2 | 22:44 |
| *** padkrish has quit IRC | 22:44 | |
| LeanReox | enikanorov2, that are regular taps for what i saw | 22:44 |
| *** zhhuabj_ has quit IRC | 22:45 | |
| LeanReox | enikanorov2, all the ports on the dead vlan are on the network node, lbaas, dhcp ... | 22:45 |
| LeanReox | enikanorov2, on the computes no ports on dead vlan | 22:45 |
| *** zhhuabj_ has joined #openstack-neutron | 22:45 | |
| *** nlahouti has joined #openstack-neutron | 22:45 | |
| enikanorov2 | LeanReox: i see. Then it looks nearly correct flow. When lbaas port is created - it's bound to a host during port creation (API call), but if ovs agent is dead on that host - binding fails and then ovs agent puts it to 4095 | 22:45 |
| *** padkrish has joined #openstack-neutron | 22:45 | |
| *** achanda has joined #openstack-neutron | 22:46 | |
| enikanorov2 | it's only recently this issue was fixed in upstream | 22:46 |
| enikanorov2 | even I'm not sure if the fix got merged, but it is on review certainly | 22:46 |
| *** nlahouti1 has joined #openstack-neutron | 22:46 | |
| enikanorov2 | i mean the issue with ovs agent trying to get information about the port that previously failed binding. | 22:47 |
| LeanReox | enikanorov2, but the same agent that puts the port in the dead vlan , is the same agent that is marked down ? | 22:47 |
| enikanorov2 | anyway with the fix there should be less issues with ports on dead vlans | 22:47 |
| LeanReox | cause the ovs agent is running on the network node, as so the lbaas ports, how can this condition be triggered ? | 22:47 |
| enikanorov2 | LeanReox: that's possible if rabbitmq is failing | 22:47 |
| *** moha_hunt has quit IRC | 22:47 | |
| enikanorov2 | e.g. ovs agent is ok, but communication is broken | 22:48 |
| LeanReox | enikanorov2, ok, this can cause half of the lbaas ports putted in to the dead vlan, and the other ports tagged correctly ? | 22:49 |
| *** diegows has joined #openstack-neutron | 22:50 | |
| enikanorov2 | it's an issue with service ports: lbaas, router and DHCP which are bound during API call. VMs port are bound via RPC call from ovs agent, the logic is slightly different so the process succeeds | 22:50 |
| *** nlahouti has quit IRC | 22:50 | |
| *** Sukhdev has quit IRC | 22:52 | |
| *** lori is now known as lori|away | 22:52 | |
| LeanReox | enikanorov2, so, if a port is maked as port_dead() , and the bound is made via API_calls() , that port in dead state will never recover till an lbaas-agent restart ? | 22:54 |
| LeanReox | recover / updated | 22:55 |
| *** sripriya has quit IRC | 22:55 | |
| enikanorov2 | LeanReox: i'm afraid it will not recover even in that case, but i may be mistaken here | 22:55 |
| enikanorov2 | but certainly not until restart | 22:55 |
| *** Divya has joined #openstack-neutron | 22:55 | |
| *** sripriya has joined #openstack-neutron | 22:55 | |
| *** echirivella has quit IRC | 22:56 | |
| LeanReox | enikanorov2, ok, so the only what that ovs_agent try to change a port from the dead_vlan to its correct tag, its to update the admin_state_up port to true on the port table on the neutron database ? | 22:56 |
| *** hitalia has joined #openstack-neutron | 22:56 | |
| LeanReox | enikanorov2, so when it begin again their polling round, will see the difference and update the port | 22:57 |
| Divya | ptoohill: Ping | 22:57 |
| enikanorov2 | LeanReox: i'm not sure. it still will stuck on fetching binding information from the server, but since binding is failed, changing admin_state_up might not help | 22:58 |
| enikanorov2 | again, i didn't heck those paths | 22:58 |
| Divya | ptoohill: Had a question about plug_vip_port in lbaas drivers | 22:58 |
| *** pece has joined #openstack-neutron | 22:58 | |
| LeanReox | enikanorov2, ok understood, so is there any way to bring back a port from a dead vlan without restarting the whole network node including openvswitch-switch process ? | 22:59 |
| enikanorov2 | LeanReox: restarting network node will not help either. I believe simplest way is to just recreate the port. | 22:59 |
| enikanorov2 | or - backport a fix to this issue :) | 23:00 |
| LeanReox | or the only mechanism i have left is to find out wich tag should have and issue an ovs-vsctl set port PortName tag=theTag | 23:00 |
| Divya | Hi i have Lbaas related question | 23:00 |
| enikanorov2 | LeanReox: that's error-prone. those tags change with ovs agent restarts, you'll have to change it each time | 23:01 |
| enikanorov2 | each time ovs restarted | 23:01 |
| enikanorov2 | LeanReox: here's the fix https://review.openstack.org/141044 | 23:01 |
| enikanorov2 | however gerrit doesn't work at the moment | 23:02 |
| ptoohill | Divya: pong | 23:02 |
| Divya | [Lbaas] When i create multiple loadbalancers on the same host, the vip port is getting the same device_id. Is this the expected behavior and wanted to understand rationale behind setting same device_id for the lbaas port | 23:02 |
| Divya | ptoohill: Hi | 23:02 |
| Divya | ptoohill: When i create multiple loadbalancers on the same host, the vip port is getting the same device_id. Is this the expected behavior and wanted to understand rationale behind setting same device_id for the lbaas port | 23:02 |
| ptoohill | join #openstack-lbaas for lbaas related topics | 23:02 |
| LeanReox | enikanorov2, can be review.openstack.org be down ? | 23:02 |
| *** aranjan has joined #openstack-neutron | 23:03 | |
| Divya | ptoohill: thanks will do that | 23:03 |
| enikanorov2 | LeanReox: yea, it's down now, I hope it will be up in few hours | 23:03 |
| *** emagana has quit IRC | 23:03 | |
| *** aranjan has quit IRC | 23:04 | |
| LeanReox | enikanorov2, i see on the conf.db openvswitch database that the tag is persistent: {"_date":1427831618555,"Port":{"5fbb01d9-4564-4386-9998-c1deb8aca7a5":{"tag":3}},"_comment":"ovs-vsctl: /usr/bin/ovs-vsctl --timeout=10 set Port tap61a741ce-f0 tag=3","Open_vSwitch":{"d0ec36b6-f81e-4ae6-94ad-3ee2a6d1940d":{"next_cfg":167}}} | 23:04 |
| *** yuanying has joined #openstack-neutron | 23:04 | |
| *** padkrish has quit IRC | 23:04 | |
| *** yamamoto has joined #openstack-neutron | 23:05 | |
| *** padkrish has joined #openstack-neutron | 23:05 | |
| enikanorov2 | LeanReox: those tags separate networks on the integration bridge. After restart ovs agent reassing local vlan tags | 23:05 |
| *** padkrish has quit IRC | 23:05 | |
| enikanorov2 | *reassigns | 23:05 |
| *** padkrish has joined #openstack-neutron | 23:06 | |
| *** aranjan has joined #openstack-neutron | 23:07 | |
| *** pece has quit IRC | 23:08 | |
| carl_baldwin | ZZelle: I was thinking again about skipping conntrack for ingress packets on the gateway interface. | 23:08 |
| LeanReox | enikanorov2, is that the way it works on icehouse ? cause in juno it just assign the same tag every time | 23:08 |
| enikanorov2 | LeanReox: i'm not sure about that. assigning same tags is possible, but you can't rely on that | 23:09 |
| enikanorov2 | sice it depends on the order in which networks and ports are created | 23:09 |
| carl_baldwin | ZZelle: Still need conntrack with shared SNAT. I think we need a rule to not NOTRACK for the router's dedicated IP address. | 23:09 |
| *** subscope has quit IRC | 23:10 | |
| enikanorov2 | if you just do several restarts - it will indeed assign same tags, just because nothing changes between restarts | 23:10 |
| *** sripriya has quit IRC | 23:11 | |
| *** cbouch has quit IRC | 23:11 | |
| *** dboik_ has quit IRC | 23:12 | |
| ZZelle | carl_baldwin, indeed | 23:12 |
| ZZelle | carl_baldwin, it seems we need to define some functional tests to ensure it behaves as expected | 23:13 |
| carl_baldwin | ZZelle: I admit, this is little more complex than I thought it would be. Yes, functional tests would be very good. | 23:13 |
| LeanReox | enikanorov2, ok i understand, but its responsability of the first start of lbaas-agent to bind a port for the first time, and if cannot bind, should resoult in a port binding failed, and then the port on a dead vlan ? | 23:15 |
| LeanReox | enikanorov2, im not sure if it is lbaas-agent or neutron-plugin-openvswitch-agent | 23:15 |
| *** britthouser has joined #openstack-neutron | 23:16 | |
| enikanorov2 | LeanReox: none of the two, i guess. vip port is a service port, which binding is done when you create a VIP | 23:16 |
| enikanorov2 | so binding may fail when you create a vip | 23:16 |
| enikanorov2 | and then not much you can do | 23:16 |
| ZZelle | carl_baldwin, i will implement it in a specific change otherwise TcNatDriver change will become tool large | 23:17 |
| carl_baldwin | ZZelle: ack | 23:17 |
| *** ivar-laz_ has joined #openstack-neutron | 23:18 | |
| ZZelle | carl_baldwin, if you have some times, could you look at https://review.openstack.org/#/c/177245/16/neutron/agent/linux/tc_lib.py and checks tc integration and docstrings? | 23:18 |
| *** nuritv has quit IRC | 23:18 | |
| *** ivar-laz_ has quit IRC | 23:19 | |
| *** ivar-laz_ has joined #openstack-neutron | 23:19 | |
| *** Longgeek_ has quit IRC | 23:20 | |
| *** carl_baldwin has quit IRC | 23:21 | |
| LeanReox | ok enikanorov2 , suppose i have to migrate from one lbaas agent to another, being the another lbaas agent on another machine, after i changed the poolloadbalanceragentbindings table with the new agent id, if a vip is a service port, if that port failed to bind to that new host on another time, this will result in that ports / that loadbalancers being put in the dead vlan | 23:21 |
| *** ivar-lazzaro has quit IRC | 23:21 | |
| enikanorov2 | i guess so. lbaas don't have built-in support for moving VIP from one host to another, unfortunately | 23:22 |
| enikanorov2 | at least, lbaas v1.0 | 23:23 |
| LeanReox | enikanorov2, ok thanks for your help, ill keep troubleshooting this to see if i can come up with a theory | 23:24 |
| *** watanabe_isao has joined #openstack-neutron | 23:27 | |
| *** arif-ali has quit IRC | 23:28 | |
| *** botang has joined #openstack-neutron | 23:29 | |
| *** arif-ali has joined #openstack-neutron | 23:29 | |
| *** dims has quit IRC | 23:34 | |
| *** moshele has quit IRC | 23:36 | |
| *** hitalia has quit IRC | 23:37 | |
| *** marun has quit IRC | 23:37 | |
| *** tidwellr has left #openstack-neutron | 23:37 | |
| *** dims has joined #openstack-neutron | 23:39 | |
| *** mturpin has quit IRC | 23:43 | |
| *** dims has quit IRC | 23:43 | |
| *** hitalia has joined #openstack-neutron | 23:44 | |
| *** yamamoto has quit IRC | 23:46 | |
| *** oomichi has joined #openstack-neutron | 23:52 | |
| -openstackstatus- NOTICE: Gerrit is going offline while we perform an emergency downgrade to version 2.8. | 23:54 | |
| *** ChanServ changes topic to "Gerrit is going offline while we perform an emergency downgrade to version 2.8." | 23:54 | |
| *** dboik has joined #openstack-neutron | 23:55 | |
| *** mchestr has quit IRC | 23:55 | |
| *** mchestr has joined #openstack-neutron | 23:55 | |
| *** ZZelle has quit IRC | 23:55 | |
| *** botang has quit IRC | 23:56 | |
| *** dboik_ has joined #openstack-neutron | 23:56 | |
| *** sdake_ has joined #openstack-neutron | 23:57 | |
| *** debedb has joined #openstack-neutron | 23:58 | |
| *** banix has joined #openstack-neutron | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!