| mtreinish | oomichi: I don't see the problem. Yes the congress tempest plugin is doing things the wrong way, but they already were before yamamoto's patch | 00:00 |
|---|---|---|
| mtreinish | all yamamoto's patch does is make the congress tempest plugin a bit easier to maintain | 00:00 |
| mtreinish | they'll still be subject to random breakages when we change the scenario tests in tempest proper | 00:00 |
| mtreinish | but they already were before that cleanup | 00:00 |
| *** zackf has quit IRC | 00:01 | |
| *** rossella_s has quit IRC | 00:01 | |
| openstackrecheck | Console logs not available after 20:08s for gate-tempest-dsvm-nnet-ubuntu-xenial 348449,6,fcd4414 | 00:13 |
| yamamoto | andreaf: is this correct understanding of your mail? https://review.openstack.org/#/c/439238/ | 00:14 |
| *** yangyapeng has quit IRC | 00:15 | |
| oomichi | yamamoto: "it switches a wrong one to a wrong but better one." humm, yeah that is correct | 00:20 |
| *** Swami has quit IRC | 00:26 | |
| *** tmatsu has joined #openstack-qa | 00:29 | |
| openstackgerrit | Jaewoo Park proposed openstack/patrole master: Users RBAC test for Keystone API v2 users https://review.openstack.org/439137 | 00:30 |
| *** rajinir_ has quit IRC | 00:31 | |
| openstackgerrit | Nicolas Helgeson proposed openstack/tempest master: Identity V3-ext Oauth1 Consumers Client https://review.openstack.org/433818 | 00:32 |
| openstackrecheck | Console logs not available after 20:04s for gate-tempest-dsvm-neutron-dvr-ubuntu-xenial 438283,1,fb13660 | 00:33 |
| *** salv-orlando has quit IRC | 00:33 | |
| *** jamielennox is now known as jamielennox|away | 00:36 | |
| *** gyee has joined #openstack-qa | 00:40 | |
| *** fnaval has quit IRC | 00:44 | |
| *** jamielennox|away is now known as jamielennox | 00:44 | |
| openstackgerrit | Merged openstack-infra/devstack-gate master: Gather /proc/cpuinfo https://review.openstack.org/433949 | 00:45 |
| *** hongbin has quit IRC | 00:46 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Add network_checker under tempest.lib https://review.openstack.org/439253 | 00:48 |
| oomichi | yamamoto: mtreinish: how about ^^^ ? | 00:49 |
| oomichi | these methods could become stable by separating tempest.conf | 00:50 |
| oomichi | https://review.openstack.org/#/c/425012 itself is good for me anyways | 00:53 |
| openstackrecheck | Console logs not available after 20:11s for gate-nova-python27-ubuntu-xenial 330285,34,37843be | 00:53 |
| openstackgerrit | zhufl proposed openstack/tempest master: Add show_server when validatable=True in create_test_server https://review.openstack.org/438829 | 00:53 |
| *** nadya has joined #openstack-qa | 00:55 | |
| *** cody-somerville has joined #openstack-qa | 00:59 | |
| *** nadya has quit IRC | 01:00 | |
| openstackgerrit | amrith proposed openstack-dev/devstack master: put mysql on a memory diet https://review.openstack.org/438668 | 01:00 |
| *** rfolco has quit IRC | 01:03 | |
| *** rfolco has joined #openstack-qa | 01:04 | |
| *** rfolco has quit IRC | 01:04 | |
| openstackrecheck | Console logs not available after 20:11s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 432787,10,bbcb7d9 | 01:13 |
| *** yangyapeng has joined #openstack-qa | 01:13 | |
| *** alex_xu has quit IRC | 01:19 | |
| *** alex_xu has joined #openstack-qa | 01:19 | |
| *** alex_xu has quit IRC | 01:25 | |
| *** alex_xu has joined #openstack-qa | 01:25 | |
| *** alex_xu has quit IRC | 01:26 | |
| *** alex_xu has joined #openstack-qa | 01:27 | |
| openstackrecheck | Console logs not available after 20:10s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 430940,14,771d4e7 | 01:33 |
| *** tmatsu_ has joined #openstack-qa | 01:38 | |
| *** cody-somerville has quit IRC | 01:38 | |
| *** tmatsu has quit IRC | 01:42 | |
| *** pvaneck_ has quit IRC | 01:43 | |
| zhufl | gmann:ping | 01:45 |
| gmann | zhufl: pong | 01:47 |
| zhufl | can you take a look at test_list_servers_filtered_by_ip? It's strange | 01:48 |
| gmann | zhufl: which one, link | 01:48 |
| zhufl | https://github.com/openstack/nova/blob/63805735c25a54ad1b9b97e05080c1a6153d8e22/nova/compute/api.py get_all | 01:48 |
| zhufl | oh, I mean the testcase test_list_servers_filtered_by_ip | 01:48 |
| gmann | zhufl: ok, i opened that | 01:50 |
| zhufl | from nova code, we can see that filter by ip is always "regexp match", but test_list_servers_filtered_by_ip seems don't think so | 01:50 |
| gmann | zhufl: yes it is regex on nova side - https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/schemas/servers.py#L355 | 01:53 |
| zhufl | gmann: so test_list_servers_filtered_by_ip seems not proper | 01:53 |
| openstackgerrit | Sean Dague proposed openstack-infra/devstack-gate master: Support setting a variable to get local.conf merged late https://review.openstack.org/438957 | 01:53 |
| openstackgerrit | Sean Dague proposed openstack-infra/devstack-gate master: bump devstack-tools to 0.4.0 https://review.openstack.org/439269 | 01:53 |
| openstackrecheck | Console logs not available after 20:12s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 434810,5,2d4ad52 | 01:54 |
| gmann | zhufl: did not get, tests trying to filter with complete IP addr | 01:54 |
| zhufl | gmann: if server1's ip is 10.1.1.1 and server2's 10.1.1.11, then test_list_servers_filtered_by_ip will fail | 01:54 |
| gmann | zhufl: and next tests with part of it as regex | 01:54 |
| gmann | zhufl: in that case it will have only server1 in list from nova right | 01:55 |
| gmann | zhufl: and if tests filter with 10.1.1 then it should have both as tested in next test case - test_list_servers_filtered_by_ip_regex | 01:56 |
| zhufl | gmann: the testcase expects only server1 is in the list, but Nova will return both in the list (10.1.1.1 can also match 10.1.1.11) | 01:56 |
| gmann | zhufl: ah i got you. | 01:56 |
| gmann | zhufl: strange combination you thought :) | 01:57 |
| zhufl | gmann: the testcase will sometimes pass and sometimes fail :) | 01:58 |
| gmann | zhufl: and that seems what nova should return. can you reproduce that? i mean you encountered that ? | 01:58 |
| zhufl | gmann: yeah | 01:58 |
| zhufl | gmann: with "regexp match", 10.1.1.1 is sure to match 10.1.1.11 | 01:59 |
| zhufl | gmann: but it will not match 10.1.1.2 | 01:59 |
| *** zackf has joined #openstack-qa | 02:00 | |
| gmann | zhufl: yea, can you file a bug. | 02:00 |
| zhufl | gmann: ok, and what's your suggestion for the sulution? | 02:01 |
| *** tmatsu has joined #openstack-qa | 02:02 | |
| gmann | zhufl: 2 options. 1. remove the tests but defcore use that 2. improve tests about check all server ips does not fall in same regex if so then do the assert checks | 02:03 |
| *** dave-mccowan has quit IRC | 02:03 | |
| gmann | i would prefer the 2nd one. | 02:03 |
| *** zackf has quit IRC | 02:04 | |
| gmann | zhufl: you can do just by doing the regex match in tests first and then do assert checks accordingly single servers or multiple in server lists from nova | 02:05 |
| zhufl | gmann: yes, and the code will become a bit longer, do you think it acceptable? :) | 02:05 |
| *** tmatsu_ has quit IRC | 02:05 | |
| gmann | zhufl: yea, that should be fine. it explicitly checks exact match of ips filter and make sure nova does works fine in those cases also | 02:07 |
| gmann | zhufl: i could have go for removal if defcore was not using it | 02:07 |
| zhufl | gmann: defcore seems to have the rule of removing testcases? | 02:08 |
| *** tmatsu_ has joined #openstack-qa | 02:09 | |
| gmann | zhufl: yea we can remove but testing with exact regex is nice | 02:10 |
| zhufl | gmann: ok, then I'll try to fix it | 02:11 |
| gmann | zhufl: yea push fix, we will see how good or bad that is | 02:12 |
| gmann | zhufl: seems like other issues in that file. ll check later | 02:13 |
| gmann | zhufl: anyways nice catch | 02:13 |
| *** tmatsu has quit IRC | 02:13 | |
| *** gvrangan has quit IRC | 02:14 | |
| openstackrecheck | Console logs not available after 20:11s for gate-dragonflow-pep8-ubuntu-xenial 424146,7,f576645 | 02:14 |
| *** gyee has quit IRC | 02:14 | |
| zhufl | gmann: there are 3 ways, which one do you prefer? | 02:15 |
| zhufl | gmann:1) create servers with specified ips, such as 1.1.1.1, 1.1.1.2, 1.1.1.3 | 02:15 |
| zhufl | 2) find the longest ip in servers, and filter it by that ip, then only one server will be returned | 02:15 |
| zhufl | 3) filter using server1's ip, and using assertIn or assertNotIn according to whether other server ip is part of server1's ip | 02:15 |
| gmann | zhufl: 2nd one is much easy and exactly tests what we want in that tests. exact match of regex | 02:19 |
| zhufl | gmann: oh, that's ok | 02:21 |
| *** tmatsu has joined #openstack-qa | 02:21 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Add network_checker under tempest.lib https://review.openstack.org/439253 | 02:22 |
| *** tmatsu_ has quit IRC | 02:24 | |
| *** mtanino has quit IRC | 02:25 | |
| openstackrecheck | Console logs not available after 20:05s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 434817,5,58ebb54 | 02:34 |
| openstackgerrit | zhufl proposed openstack/tempest master: Add show_server when validatable=True in create_test_server https://review.openstack.org/438829 | 02:36 |
| *** Apoorva_ has joined #openstack-qa | 02:36 | |
| *** Apoorva has quit IRC | 02:39 | |
| *** Apoorva_ has quit IRC | 02:40 | |
| *** gvrangan has joined #openstack-qa | 02:44 | |
| *** ramishra has joined #openstack-qa | 02:49 | |
| *** sapcc-bot1 has joined #openstack-qa | 02:52 | |
| *** carthaca_ has quit IRC | 02:52 | |
| *** sapcc-bot has quit IRC | 02:52 | |
| *** carthaca_ has joined #openstack-qa | 02:52 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/tempest master: Keep scenario manager temporary copy for plugins https://review.openstack.org/439291 | 02:52 |
| gmann | mtreinish andreaf oomichi masayukig afazekas sdague jordanP : can you look into this - https://review.openstack.org/#/c/439291/1 | 02:54 |
| openstackrecheck | Console logs not available after 20:22s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 434809,5,6011e56 | 02:54 |
| gmann | another way to solve the scenario refactoring work. | 02:54 |
| *** nadya has joined #openstack-qa | 02:56 | |
| *** nadya has quit IRC | 03:00 | |
| *** ihrachys has joined #openstack-qa | 03:03 | |
| *** gvrangan has quit IRC | 03:03 | |
| *** yamahata has quit IRC | 03:08 | |
| *** zackf has joined #openstack-qa | 03:12 | |
| *** ihrachys has quit IRC | 03:13 | |
| openstackrecheck | Console logs not available after 20:07s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 435941,4,bc24fca | 03:14 |
| *** rfolco has joined #openstack-qa | 03:24 | |
| *** rfolco has quit IRC | 03:24 | |
| *** zhufl has quit IRC | 03:28 | |
| *** zhufl has joined #openstack-qa | 03:29 | |
| openstackgerrit | Felipe Monteiro proposed openstack/patrole master: Compute API Create Backup Test https://review.openstack.org/438676 | 03:31 |
| *** huanxie has joined #openstack-qa | 03:33 | |
| *** salv-orlando has joined #openstack-qa | 03:34 | |
| openstackrecheck | Console logs not available after 20:10s for gate-tacker-pep8-ubuntu-xenial 438211,2,2864297 | 03:35 |
| *** amotoki_ has joined #openstack-qa | 03:36 | |
| *** salv-orlando has quit IRC | 03:39 | |
| *** amotoki has quit IRC | 03:39 | |
| *** armax has joined #openstack-qa | 03:45 | |
| openstackrecheck | Console logs not available after 20:11s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 434431,5,c2c4a83 | 03:55 |
| *** cody-somerville has joined #openstack-qa | 03:59 | |
| *** armax has quit IRC | 04:11 | |
| *** shausy has joined #openstack-qa | 04:12 | |
| *** ihrachys has joined #openstack-qa | 04:13 | |
| *** ihrachys has quit IRC | 04:15 | |
| openstackrecheck | Console logs not available after 20:08s for gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial 434453,7,dadbf26 | 04:15 |
| *** gvrangan has joined #openstack-qa | 04:22 | |
| *** ramishra has quit IRC | 04:26 | |
| *** ramishra has joined #openstack-qa | 04:27 | |
| openstackgerrit | Jianghua Wang proposed openstack-dev/devstack master: WIP: Support generate the BASE Ubuntu VM XVA. https://review.openstack.org/439380 | 04:29 |
| openstackrecheck | Console logs not available after 20:10s for gate-functional-dsvm-magnum-k8s-ubuntu-trusty 437069,5,97cc295 | 04:35 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add migration to add common metadata fields to runs table https://review.openstack.org/334556 | 04:39 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add a new db manage command to migrate run_metadata https://review.openstack.org/334557 | 04:40 |
| openstackgerrit | fumihiko kakuma proposed openstack-dev/devstack master: DO NOT MERGE: This is to check ryu master https://review.openstack.org/439383 | 04:44 |
| *** yamahata has joined #openstack-qa | 04:45 | |
| *** huanxie has quit IRC | 04:48 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add new graph time for variance of individual tests https://review.openstack.org/230611 | 04:50 |
| *** nadya has joined #openstack-qa | 04:54 | |
| openstackrecheck | Console logs not available after 20:08s for gate-keystone-python27-ubuntu-xenial 438206,2,5f7cf11 | 04:55 |
| *** nadya has quit IRC | 04:56 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add new graph time for variance of individual tests https://review.openstack.org/230611 | 04:59 |
| *** amotoki has joined #openstack-qa | 05:07 | |
| *** amotoki_ has quit IRC | 05:09 | |
| openstackrecheck | Console logs not available after 20:08s for gate-keystone-pep8-ubuntu-xenial 438207,2,f71776b | 05:15 |
| *** carthaca_ has quit IRC | 05:16 | |
| *** carthaca_1 has joined #openstack-qa | 05:16 | |
| *** artom has quit IRC | 05:17 | |
| *** artom_ has joined #openstack-qa | 05:17 | |
| *** agurenko has joined #openstack-qa | 05:23 | |
| *** Alex_Staf has joined #openstack-qa | 05:26 | |
| *** agurenko has quit IRC | 05:29 | |
| *** agurenko has joined #openstack-qa | 05:30 | |
| *** gvrangan has quit IRC | 05:35 | |
| *** salv-orlando has joined #openstack-qa | 05:35 | |
| openstackrecheck | Console logs not available after 20:30s for gate-fuxi-dsvm-fullstack 420386,3,c5ceb41 | 05:36 |
| *** sandanar has joined #openstack-qa | 05:36 | |
| *** e0ne has joined #openstack-qa | 05:36 | |
| *** e0ne has quit IRC | 05:38 | |
| *** gouthamr has quit IRC | 05:40 | |
| openstackgerrit | zhufl proposed openstack/tempest master: Add show_server when validatable=True in create_test_server https://review.openstack.org/438829 | 05:40 |
| *** salv-orlando has quit IRC | 05:40 | |
| openstackgerrit | fumihiko kakuma proposed openstack-dev/devstack master: DO NOT MERGE: This is to check ryu master https://review.openstack.org/439383 | 05:44 |
| *** sridharg has joined #openstack-qa | 05:45 | |
| *** cody-somerville has quit IRC | 05:49 | |
| openstackrecheck | Console logs not available after 20:10s for gate-dragonflow-python27-ubuntu-xenial 433609,3,717daeb | 05:56 |
| *** edannon has joined #openstack-qa | 05:56 | |
| *** gcheresh has joined #openstack-qa | 05:59 | |
| afazekas | strange approach I'll read the thread later | 06:01 |
| *** pgadiya has joined #openstack-qa | 06:16 | |
| openstackrecheck | Console logs not available after 20:08s for gate-dragonflow-python27-ubuntu-xenial 429666,4,ffeb778 | 06:16 |
| *** DotanSoko has joined #openstack-qa | 06:19 | |
| *** salv-orlando has joined #openstack-qa | 06:26 | |
| openstackgerrit | Vasyl Saienko proposed openstack-infra/devstack-gate master: DNR: add enable_plugin ironic to subnode local.conf https://review.openstack.org/439044 | 06:27 |
| *** DotanSoko_ has joined #openstack-qa | 06:30 | |
| *** DotanSoko_ has quit IRC | 06:30 | |
| *** DotanSoko_ has joined #openstack-qa | 06:32 | |
| *** DotanSoko has quit IRC | 06:33 | |
| openstackrecheck | Console logs not available after 20:09s for gate-tempest-dsvm-nnet-ubuntu-xenial 348449,6,4c4e9ef | 06:36 |
| *** sshnaidm|off is now known as sshnaidm | 06:49 | |
| *** sridharg has quit IRC | 06:50 | |
| *** rcernin has joined #openstack-qa | 06:51 | |
| *** sigmavirus has quit IRC | 06:52 | |
| *** salv-orl_ has joined #openstack-qa | 06:52 | |
| *** sigmavirus has joined #openstack-qa | 06:53 | |
| *** sigmavirus is now known as Guest53153 | 06:54 | |
| *** salv-orlando has quit IRC | 06:55 | |
| openstackrecheck | Console logs not available after 20:11s for gate-fuxi-dsvm-fullstack 438292,2,fa37f09 | 06:57 |
| *** sridharg has joined #openstack-qa | 07:02 | |
| *** yamamoto has quit IRC | 07:07 | |
| *** tesseract has joined #openstack-qa | 07:12 | |
| openstackgerrit | Vasyl Saienko proposed openstack-infra/devstack-gate master: DNR: add enable_plugin ironic to subnode local.conf https://review.openstack.org/439044 | 07:14 |
| openstackrecheck | Console logs not available after 20:07s for gate-neutron-lbaasv2-dsvm-api-ubuntu-xenial 423543,28,e502416 | 07:17 |
| *** amotoki has quit IRC | 07:20 | |
| *** salv-orl_ has quit IRC | 07:25 | |
| *** itzikb has joined #openstack-qa | 07:27 | |
| openstackgerrit | zhufl proposed openstack/tempest master: Correct test_list_servers_filtered_by_ip https://review.openstack.org/439428 | 07:32 |
| *** amotoki has joined #openstack-qa | 07:34 | |
| *** tmatsu_ has joined #openstack-qa | 07:35 | |
| *** DotanSoko has joined #openstack-qa | 07:35 | |
| openstackgerrit | zhufl proposed openstack/tempest master: Correct test_list_servers_filtered_by_ip https://review.openstack.org/439428 | 07:35 |
| openstackrecheck | Console logs not available after 20:08s for gate-dragonflow-python27-ubuntu-xenial 424145,7,86281d9 | 07:37 |
| *** DotanSoko_ has quit IRC | 07:38 | |
| *** tmatsu has quit IRC | 07:38 | |
| *** tmatsu has joined #openstack-qa | 07:38 | |
| *** yamahata has quit IRC | 07:38 | |
| *** yamamoto has joined #openstack-qa | 07:40 | |
| *** tmatsu_ has quit IRC | 07:41 | |
| openstackgerrit | zhufl proposed openstack/tempest master: Correct test_list_servers_filtered_by_ip https://review.openstack.org/439428 | 07:42 |
| *** nadya has joined #openstack-qa | 07:42 | |
| *** amotoki has quit IRC | 07:42 | |
| *** dimtruck is now known as zz_dimtruck | 07:48 | |
| openstackrecheck | Console logs not available after 20:08s for gate-fuxi-dsvm-fullstack 438292,3,5d10317 | 07:57 |
| *** markus_z has joined #openstack-qa | 07:57 | |
| *** abehl has joined #openstack-qa | 07:58 | |
| openstackgerrit | Béla Vancsics proposed openstack/tempest master: Reduced the complexity https://review.openstack.org/437871 | 07:58 |
| *** sshwarts has joined #openstack-qa | 08:00 | |
| *** tmatsu_ has joined #openstack-qa | 08:05 | |
| *** tmatsu has quit IRC | 08:08 | |
| *** tmatsu_ has quit IRC | 08:10 | |
| *** tmatsu has joined #openstack-qa | 08:10 | |
| *** dfflanders has quit IRC | 08:12 | |
| *** tmatsu_ has joined #openstack-qa | 08:14 | |
| *** tmatsu has quit IRC | 08:16 | |
| openstackrecheck | Console logs not available after 20:12s for gate-blazar-pep8-ubuntu-xenial 438293,1,f7d7ed5 | 08:17 |
| *** shardy has joined #openstack-qa | 08:19 | |
| *** tmatsu has joined #openstack-qa | 08:21 | |
| openstackgerrit | Béla Vancsics proposed openstack/tempest master: Reduced the complexity https://review.openstack.org/437871 | 08:22 |
| *** tmatsu__ has joined #openstack-qa | 08:23 | |
| *** salv-orlando has joined #openstack-qa | 08:24 | |
| *** tmatsu_ has quit IRC | 08:24 | |
| *** sridharg has quit IRC | 08:25 | |
| *** sridharg has joined #openstack-qa | 08:25 | |
| *** sridharg has quit IRC | 08:25 | |
| *** tmatsu has quit IRC | 08:26 | |
| *** sridharg has joined #openstack-qa | 08:27 | |
| *** amotoki has joined #openstack-qa | 08:32 | |
| *** tmatsu has joined #openstack-qa | 08:32 | |
| *** abehl has quit IRC | 08:33 | |
| *** abehl has joined #openstack-qa | 08:33 | |
| *** tmatsu_ has joined #openstack-qa | 08:33 | |
| *** tmatsu_ has quit IRC | 08:35 | |
| *** tmatsu__ has quit IRC | 08:36 | |
| *** tmatsu has quit IRC | 08:37 | |
| openstackrecheck | Console logs not available after 20:08s for gate-neutron-lbaasv2-dsvm-api-ubuntu-xenial 429398,12,4c5a0e1 | 08:37 |
| openstackrecheck | Console logs not available after 20:08s for gate-functional-dsvm-magnum-swarm-ubuntu-xenial 436559,2,f80f776 | 08:58 |
| *** amoralej|off is now known as amoralej | 08:58 | |
| *** tmatsu has joined #openstack-qa | 09:00 | |
| *** alex_xu has quit IRC | 09:01 | |
| *** alex_xu has joined #openstack-qa | 09:02 | |
| *** martinkopec has joined #openstack-qa | 09:02 | |
| *** rossella_s has joined #openstack-qa | 09:03 | |
| *** pgadiya has quit IRC | 09:04 | |
| *** pgadiya has joined #openstack-qa | 09:05 | |
| openstackgerrit | zhufl proposed openstack/tempest master: Remove unused client in setup_clients https://review.openstack.org/439473 | 09:06 |
| *** dtantsur|afk is now known as dtantsur | 09:07 | |
| *** tosky has joined #openstack-qa | 09:11 | |
| *** tmatsu has quit IRC | 09:18 | |
| openstackrecheck | Console logs not available after 20:17s for gate-rally-dsvm-cli 438299,1,956fdbf | 09:18 |
| andreaf | gmann: well the fact that we keep a copy of that file there does not mean we won't break plugins | 09:19 |
| openstackgerrit | zhufl proposed openstack/tempest master: Remove unused variable and unnecessary instance variable https://review.openstack.org/439481 | 09:19 |
| andreaf | gmann: the moment we stop testing that base class we may break it anytime | 09:19 |
| *** k4n0 has joined #openstack-qa | 09:19 | |
| openstackgerrit | Oksana Voshchana proposed openstack-infra/devstack-gate master: WIP: Test SSL wsgi https://review.openstack.org/414986 | 09:19 |
| andreaf | gmann: which means we will hold a piece of untested and bit rotting code in there | 09:20 |
| andreaf | gmann: I fear that will just postpone the pain... I'd rather have it the other way around - even though that's not a perfect solution either | 09:20 |
| gmann | andreaf: we will not make any change in that copy of manager. and yea it will be untested from tempest side but that is temporary copy which will be removed soon | 09:21 |
| andreaf | gmann: I don't think there's a really good solution now, but we definitely need speed in that refactor, we have other urgent things to work on and I would like us to avoid spending time in endless cycles for the refector | 09:21 |
| andreaf | yeah but because it's not tested it may break - it depends on other parts of tempest as well.... | 09:22 |
| andreaf | that of course can happen the same for that code on plugin side | 09:22 |
| gmann | andreaf: actually IMO if we ask plugins to switch to other interface we should give them helpers or stable interface, otherwise they switch or change code multiple times | 09:22 |
| gmann | andreaf: yea. | 09:22 |
| *** gfidente|afk is now known as gfidente | 09:23 | |
| openstackgerrit | Oksana Voshchana proposed openstack-infra/devstack-gate master: WIP: Test SSL wsgi https://review.openstack.org/414986 | 09:23 |
| gmann | andreaf: if we can give them a single notification to change their code to ready interface instead of asking them to keep local copy of interface or other way of doing that and later once we are ready with helper ask them to change or use those helper | 09:24 |
| *** yamamoto has quit IRC | 09:24 | |
| gmann | andreaf: yea +1 on speedy the refactor work and give less pain to plugins also. i thought keeping copy in tempest can do that | 09:25 |
| andreaf | gmann: well I don't think scenario is a stable interface really - the test base class will be - maybe we will make some helpers as a stable interface | 09:25 |
| gmann | andreaf: yea same thing like common function as helper like connectivity checks etc | 09:25 |
| andreaf | gmann: probably the most interesting bit of scenario base class is the credentials bit which comes from test.py | 09:26 |
| *** tosky_ has joined #openstack-qa | 09:26 | |
| andreaf | gmann: so for me we would not be asking people of switching twice, just host the base class locally and I guess teams may want to prune it to what they really need | 09:27 |
| *** tosky has quit IRC | 09:27 | |
| *** lucas-afk is now known as lucasagomes | 09:28 | |
| openstackgerrit | Merged openstack-dev/devstack stable/ocata: Fix api extension settings in tempest for Ocata https://review.openstack.org/438726 | 09:28 |
| * andreaf afk - new fridge delivery - I'll be back as soon as possible | 09:29 | |
| gmann | andreaf: but that needs each plugins to keep the copy and then if we have stable helper for them it will be less encouraging for them to move to those | 09:29 |
| openstackgerrit | Merged openstack-dev/devstack master: lib/neutron: Fix conf handling https://review.openstack.org/437083 | 09:29 |
| openstackgerrit | Merged openstack-dev/devstack master: lib/neutron: Fix an extra comma in service_plugins https://review.openstack.org/437084 | 09:29 |
| openstackgerrit | Merged openstack-dev/devstack master: nova: Fix comments for _config_nova_apache_wsgi https://review.openstack.org/438823 | 09:29 |
| openstackrecheck | Console logs not available after 20:08s for gate-rally-dsvm-cli 438275,2,50275e1 | 09:38 |
| openstackgerrit | Oksana Voshchana proposed openstack-infra/devstack-gate master: WIP: Test SSL wsgi https://review.openstack.org/414986 | 09:42 |
| openstackgerrit | Béla Vancsics proposed openstack/tempest master: Remove unused parameter https://review.openstack.org/439499 | 09:45 |
| openstackgerrit | Oksana Voshchana proposed openstack-infra/devstack-gate master: WIP: Test SSL wsgi https://review.openstack.org/414986 | 09:50 |
| *** carthaca_1 has quit IRC | 09:51 | |
| *** sapcc-bot1 has quit IRC | 09:51 | |
| *** databus23_ has joined #openstack-qa | 09:52 | |
| *** carthaca_ has joined #openstack-qa | 09:52 | |
| *** dgonzalez_ has joined #openstack-qa | 09:52 | |
| *** tpatzig_ has joined #openstack-qa | 09:52 | |
| *** david_1 has joined #openstack-qa | 09:52 | |
| *** databus23_ has quit IRC | 09:54 | |
| *** tpatzig_ has quit IRC | 09:54 | |
| *** dgonzalez_ has quit IRC | 09:54 | |
| *** david_1 has quit IRC | 09:54 | |
| *** sapcc-bot has joined #openstack-qa | 09:54 | |
| *** carthaca_ has quit IRC | 09:57 | |
| *** sapcc-bot has quit IRC | 09:57 | |
| *** david_1 has joined #openstack-qa | 09:57 | |
| *** databus23_ has joined #openstack-qa | 09:57 | |
| *** carthaca_ has joined #openstack-qa | 09:57 | |
| *** sapcc-bot has joined #openstack-qa | 09:57 | |
| *** dgonzalez_ has joined #openstack-qa | 09:57 | |
| *** tpatzig_ has joined #openstack-qa | 09:57 | |
| openstackrecheck | Console logs not available after 20:07s for gate-tempest-dsvm-murano-api-ubuntu-xenial 437818,1,a251f84 | 09:58 |
| *** david_1 has quit IRC | 09:59 | |
| *** databus23_ has quit IRC | 09:59 | |
| *** dgonzalez_ has quit IRC | 09:59 | |
| *** tpatzig_ has quit IRC | 09:59 | |
| *** yangyapeng has quit IRC | 09:59 | |
| *** nadya has quit IRC | 10:01 | |
| *** penchal has joined #openstack-qa | 10:02 | |
| openstackgerrit | Béla Vancsics proposed openstack/tempest master: Remove unused variable https://review.openstack.org/439499 | 10:09 |
| openstackgerrit | Merged openstack/tempest master: Updated from global requirements https://review.openstack.org/432114 | 10:10 |
| *** e0ne has joined #openstack-qa | 10:10 | |
| *** mvk has quit IRC | 10:12 | |
| penchal | hi when i am running designate tempest plugin along with tempest getting ImportError: No module named dns | 10:13 |
| *** dsariel has quit IRC | 10:13 | |
| penchal | anyone has any idea | 10:13 |
| penchal | i have added service type dns in keystone and provided proper keystone details in tempest.conf | 10:14 |
| openstackrecheck | Console logs not available after 20:10s for gate-group-based-policy-python27-ubuntu-xenial 426437,33,deebc09 | 10:18 |
| *** dtantsur is now known as dtantsur|bbl | 10:20 | |
| *** yamamoto has joined #openstack-qa | 10:21 | |
| openstackgerrit | fumihiko kakuma proposed openstack-dev/devstack master: DO NOT MERGE: This is to check ryu master https://review.openstack.org/439383 | 10:24 |
| *** pcaruana has joined #openstack-qa | 10:24 | |
| *** yamamoto has quit IRC | 10:25 | |
| *** penchal has quit IRC | 10:29 | |
| *** cdent has joined #openstack-qa | 10:30 | |
| *** sridharg has quit IRC | 10:30 | |
| *** alex_xu has quit IRC | 10:31 | |
| *** dsariel has joined #openstack-qa | 10:32 | |
| *** alex_xu has joined #openstack-qa | 10:33 | |
| openstackrecheck | Console logs not available after 20:11s for gate-neutron-lbaasv2-dsvm-api-ubuntu-xenial 435932,1,4963aae | 10:38 |
| *** mvk has joined #openstack-qa | 10:44 | |
| *** mvk has quit IRC | 10:45 | |
| *** alex_xu has quit IRC | 10:45 | |
| *** mvk has joined #openstack-qa | 10:45 | |
| *** alex_xu has joined #openstack-qa | 10:46 | |
| *** nadya has joined #openstack-qa | 10:49 | |
| *** markus_z has quit IRC | 10:52 | |
| *** tosky_ is now known as tosky | 10:58 | |
| openstackrecheck | Console logs not available after 20:10s for gate-tempest-dsvm-networking-odl-beryllium-snapshot-v2driver 435813,13,c78a269 | 10:59 |
| openstackgerrit | Vasyl Saienko proposed openstack-dev/grenade master: WIP: DNR https://review.openstack.org/438955 | 10:59 |
| *** erlon has joined #openstack-qa | 10:59 | |
| openstackgerrit | Sravan Kumar proposed openstack/patrole master: Compute API Server Actions Test https://review.openstack.org/438827 | 11:01 |
| *** sdague has joined #openstack-qa | 11:01 | |
| *** cdent has quit IRC | 11:05 | |
| *** amotoki has quit IRC | 11:07 | |
| openstackgerrit | Oksana Voshchana proposed openstack-infra/devstack-gate master: WIP: Test SSL wsgi https://review.openstack.org/414986 | 11:10 |
| *** dsariel has quit IRC | 11:12 | |
| *** amotoki has joined #openstack-qa | 11:13 | |
| *** Serlex has joined #openstack-qa | 11:17 | |
| openstackrecheck | Console logs not available after 20:11s for gate-tempest-docs-ubuntu-xenial 432114,21,c5e19ea | 11:19 |
| *** yamamoto has joined #openstack-qa | 11:26 | |
| *** tmatsu has joined #openstack-qa | 11:29 | |
| openstackgerrit | Oksana Voshchana proposed openstack-infra/devstack-gate master: WIP: Test SSL wsgi https://review.openstack.org/414986 | 11:30 |
| *** yamamoto has quit IRC | 11:32 | |
| *** jordanP has joined #openstack-qa | 11:32 | |
| *** mvk has quit IRC | 11:32 | |
| *** amotoki has quit IRC | 11:36 | |
| *** dtantsur|bbl is now known as dtantsur | 11:37 | |
| *** yamamoto has joined #openstack-qa | 11:38 | |
| openstackrecheck | Console logs not available after 20:14s for gate-nova-python27-ubuntu-xenial 330285,35,383b170 | 11:39 |
| *** itzikb_ has joined #openstack-qa | 11:39 | |
| *** itzikb has quit IRC | 11:39 | |
| *** yamamoto has quit IRC | 11:42 | |
| *** mvk has joined #openstack-qa | 11:45 | |
| *** yamamoto has joined #openstack-qa | 11:47 | |
| *** yamamoto has quit IRC | 11:48 | |
| *** sandanar has quit IRC | 11:54 | |
| openstackrecheck | Console logs not available after 20:13s for gate-tempest-dsvm-nnet-ubuntu-xenial 335358,48,5e2323e | 11:59 |
| *** dsariel has joined #openstack-qa | 12:00 | |
| openstackgerrit | Vasyl Saienko proposed openstack-dev/grenade master: WIP: DNR https://review.openstack.org/438955 | 12:02 |
| *** Guest53153 is now known as sigmavirus | 12:10 | |
| *** sigmavirus has quit IRC | 12:10 | |
| *** sigmavirus has joined #openstack-qa | 12:10 | |
| *** gszasz has joined #openstack-qa | 12:11 | |
| *** dave-mccowan has joined #openstack-qa | 12:14 | |
| *** markus_z has joined #openstack-qa | 12:15 | |
| *** sridharg has joined #openstack-qa | 12:17 | |
| *** lucasagomes is now known as lucas-hungry | 12:18 | |
| openstackgerrit | Kevin Benton proposed openstack-infra/elastic-recheck master: Add query for bug 1668958 https://review.openstack.org/439562 | 12:18 |
| openstack | bug 1668958 in OpenStack Compute (nova) "metadata service occasionally not returning keys" [Undecided,New] https://launchpad.net/bugs/1668958 | 12:18 |
| *** aysyd has joined #openstack-qa | 12:18 | |
| *** alex_xu has quit IRC | 12:18 | |
| *** sridharg has quit IRC | 12:19 | |
| openstackrecheck | Console logs not available after 20:13s for gate-tempest-dsvm-neutron-full-ubuntu-xenial 437511,1,842183f | 12:20 |
| *** shardy is now known as shardy_lunch | 12:20 | |
| *** sridharg has joined #openstack-qa | 12:20 | |
| *** amotoki has joined #openstack-qa | 12:21 | |
| *** alex_xu has joined #openstack-qa | 12:25 | |
| *** pgadiya has quit IRC | 12:30 | |
| *** abehl has quit IRC | 12:30 | |
| *** k4n0 has quit IRC | 12:31 | |
| openstackrecheck | Console logs not available after 20:16s for gate-group-based-policy-python27-ubuntu-xenial 426437,34,16bee8f | 12:40 |
| *** armax has joined #openstack-qa | 12:48 | |
| *** yamamoto has joined #openstack-qa | 12:49 | |
| *** dave-mccowan has quit IRC | 12:50 | |
| *** salv-orl_ has joined #openstack-qa | 12:53 | |
| *** yamamoto has quit IRC | 12:53 | |
| *** salv-orlando has quit IRC | 12:56 | |
| *** k4n0 has joined #openstack-qa | 12:56 | |
| *** gcheresh has quit IRC | 12:59 | |
| *** gcheresh_ has joined #openstack-qa | 12:59 | |
| openstackrecheck | Console logs not available after 20:13s for gate-tempest-dsvm-neutron-full-ubuntu-xenial 436117,1,d1709ab | 13:00 |
| *** amotoki has quit IRC | 13:00 | |
| *** lucas-hungry is now known as lucasagomes | 13:01 | |
| openstackgerrit | Vasyl Saienko proposed openstack-dev/grenade master: Apply OS_TEST_TIMEOUT when running tempest https://review.openstack.org/438955 | 13:03 |
| openstackgerrit | Vasyl Saienko proposed openstack-dev/grenade master: Apply OS_TEST_TIMEOUT when running tempest https://review.openstack.org/438955 | 13:03 |
| *** salv-orl_ has quit IRC | 13:08 | |
| *** grafuls has quit IRC | 13:11 | |
| *** nadya has quit IRC | 13:17 | |
| frickler | if I install tempest into a venv, running "$venv/bin/tempest run" fails with various "module not found" errors. If I instead do "source $venv/bin/activate; tempest run", everything works fine. is this the expected behaviour, a bug, or do I need to do something different? | 13:20 |
| openstackrecheck | Console logs not available after 20:05s for gate-trove-scenario-dsvm-mysql-multi-ubuntu-xenial 432120,4,f89dd87 | 13:20 |
| *** shardy_lunch is now known as shardy | 13:21 | |
| *** itzikb has joined #openstack-qa | 13:23 | |
| *** itzikb_ has quit IRC | 13:23 | |
| *** itzikb has quit IRC | 13:23 | |
| *** abehl has joined #openstack-qa | 13:24 | |
| *** itzikb has joined #openstack-qa | 13:25 | |
| *** amoralej is now known as amoralej|lunch | 13:25 | |
| *** mvk has quit IRC | 13:29 | |
| *** k4n0 has quit IRC | 13:31 | |
| *** gouthamr has joined #openstack-qa | 13:40 | |
| openstackrecheck | Console logs not available after 20:09s for gate-rally-dsvm-cli 432135,3,cd90de4 | 13:40 |
| andreaf | sdague: heh I know you're unhappy with the current credentials magic, I made a few attempts to switch that to actual fixtures... | 13:42 |
| andreaf | sdague: it's totally possible today to get credentials in a more explicit way instead of using the class level list | 13:43 |
| andreaf | sdague: but perhaps a blog post + extra docs is part of the solution? | 13:43 |
| tosky | extra_docs++ | 13:44 |
| sdague | andreaf: yeh, my concern is honestly that I'm not convinced that anyone knows (and definitely not new / medium contributors), without reading tempest logs, the complete set of API calls that any particular test is making | 13:44 |
| sdague | which means, there is no double bookkeeping of "the following 32 actions should happen" and whether they do or not, and have the right results. | 13:45 |
| sdague | and if you don't know easily in advance every single thing that a test should be doing, then it's not testing an assertion, it's just returning a success result, which may or may not be the assertion you were trying for. | 13:46 |
| andreaf | sdague: sure, but for most tests having a set of working credentials is just really setup to get to the starting point of the tests, which was the reason to make it as simple to use as possible | 13:47 |
| andreaf | sdague: but it obviously failed in being simple :D | 13:47 |
| sdague | sure, but then do that with a fixture | 13:47 |
| andreaf | sdague: Tempest relies a lot on class level fixtures i.e. a bunch of tests have a shared bit of setup | 13:48 |
| sdague | self.useFixture(project="demo", user="demo", roles="is_admin") | 13:48 |
| sdague | andreaf: right, and I actually think that's an anti pattern | 13:48 |
| sdague | that's what makes making the codebase better hard | 13:48 |
| andreaf | sdague: which is good to reduce the amount of setup to be done, but it has a number of nasty side-effects | 13:48 |
| *** databus23_ has joined #openstack-qa | 13:49 | |
| *** tpatzig_ has joined #openstack-qa | 13:49 | |
| *** dgonzalez_ has joined #openstack-qa | 13:49 | |
| *** tommy_ has joined #openstack-qa | 13:49 | |
| sdague | been slogging through similar things in Nova for the past couple of years | 13:49 |
| *** tommy_ is now known as Guest89971 | 13:49 | |
| andreaf | sdague: but it's now very hard to change that, I don't think we would have the bandwidth to move away from that | 13:49 |
| sdague | "here is a real bug fix. Oh, 120 coupled tests with shared setup now break" | 13:49 |
| *** zz_dimtruck is now known as dimtruck | 13:49 | |
| andreaf | if we had test resoruces only we could use some smart scheduler to make sure we balance resource allocation through a test run | 13:49 |
| andreaf | test level resources | 13:50 |
| sdague | andreaf: no no no no no | 13:50 |
| sdague | that's just deciding to solve it with more magic | 13:50 |
| sdague | which means the ability to assert what is going on in advance gets even less | 13:50 |
| *** dgonzalez_ has quit IRC | 13:50 | |
| *** databus23_ has quit IRC | 13:50 | |
| *** Guest89971 has quit IRC | 13:50 | |
| *** tpatzig_ has quit IRC | 13:50 | |
| andreaf | sdague: heh yeah I meant if each test declared what they need we could manage the parallel execution better | 13:51 |
| andreaf | sdague: I agree creating resources at a random time would be more magic | 13:51 |
| sdague | andreaf: no, even parallel | 13:51 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack-infra/devstack-gate master: DNM test disabled ironic *_ssh drivers https://review.openstack.org/439601 | 13:51 |
| andreaf | so maybe we just switch to using pre-provisioned credentials | 13:51 |
| andreaf | and everything else is created / destroyed by each test | 13:52 |
| sdague | because then all you are doing is performance tuning to a very specific environment | 13:52 |
| *** dustins has joined #openstack-qa | 13:52 | |
| sdague | which again reduces repeatability | 13:52 |
| sdague | because someone manually tuned variables to just barely work well on the constrained test environments of the gate | 13:52 |
| andreaf | well I think there is a balance to strike here - I agree you need very good repeatability for gate tests, but I think we had a rather good degree of that for a while | 13:54 |
| sdague | andreaf: I disagree | 13:54 |
| andreaf | and parallel execution exposed issues in the past, so if you just drop to serial you would actually lose good testing | 13:54 |
| sdague | it's fine to run parallel on light resource tests | 13:55 |
| sdague | though, honestly, it would also be fine to push that back to the projects | 13:55 |
| andreaf | "light" is relative to your test environments I guess | 13:55 |
| sdague | I think the crux of missing common ground at least includes the following: | 13:55 |
| sdague | 1) the assumption that the state of testing in projects 4 years ago, and who's responsibility that is hasn't changed over the 4 years (I think it has) | 13:56 |
| sdague | 2) the assumption that things only recently got bad for gate repeatability, vs. it's been bad for 3+ years | 13:57 |
| sdague | and instead of just backing off, we tried to build tools around it - elastic recheck, logstash, openstack-health | 13:57 |
| *** sridharg has quit IRC | 13:58 | |
| sdague | a ton of effort and manpower is going into tools that only need to exist because the repeatability is so low because we are running our test nodes at load average 10 within 100 MB of triggering an OOM kill | 13:58 |
| *** yamamoto has joined #openstack-qa | 13:58 | |
| sdague | and while there is a lot of cleverness in those tools, I definitely think it's the wrong trade off | 13:59 |
| sdague | because maintaining those tools and extending them doesn't come for free | 13:59 |
| sdague | and further splits the effort we have | 13:59 |
| *** amotoki has joined #openstack-qa | 14:00 | |
| sdague | anyway... | 14:00 |
| openstackrecheck | Console logs not available after 20:06s for gate-tempest-dsvm-networking-odl-beryllium-snapshot-v2driver 356839,6,d2305ab | 14:00 |
| jordanP | <sdague> 2) the assumption that things only recently got bad for gate repeatability, vs. it's been bad for 3+ years | 14:01 |
| jordanP | I disagree | 14:01 |
| jordanP | erf, brb in 5min | 14:01 |
| andreaf | sdague: I don't have numbers to back this, but I think in a system as complex as openstack 100% repeatability is not a real thing - and I think you get a lot of value in terms of bug discovery in being a bit borderline | 14:02 |
| sdague | andreaf: you do if it's not at the cost of blocking perfectly valid patches | 14:02 |
| andreaf | sdague: chasing things which are only related to system overload is not useful though | 14:02 |
| sdague | and you have enough people to look into those bugs | 14:02 |
| andreaf | sdague: yes I agree on not blocking people work 100% | 14:03 |
| andreaf | sdague: so we have a problem at hand and I appreciate your going back and rethinking how we got here and which effort was worth and which not | 14:03 |
| sdague | andreaf: any patch that gets to the gate should not fail landing in the gate unless there is a real code error caused by a merge gone bad | 14:03 |
| andreaf | sdague: it's all good thinking, so I've got a lot to think about now | 14:04 |
| *** itzikb has quit IRC | 14:04 | |
| *** yamamoto has quit IRC | 14:04 | |
| andreaf | sdague: yes but I think there can be a tolerance there | 14:04 |
| *** itzikb has joined #openstack-qa | 14:05 | |
| *** dustins has quit IRC | 14:05 | |
| andreaf | sdague: it the false negative rate raises too much you start to scare off developers | 14:05 |
| sdague | andreaf: you keep using future tense :) | 14:05 |
| sdague | what I'm trying to say is all that happened years ago | 14:05 |
| andreaf | and in a downstream CI the tolerance is higher as engineers "have to" work on patches | 14:06 |
| andreaf | sdague: heh ok that's just my lack english skills :D | 14:06 |
| sdague | and that's my problem with what I hear in the conversation, the assumption that it's all fine, and we need to be cautious about future impacts, instead of realizing that part of the reason it's hard to get people engaged on these issues is because they already disconnected long ago. | 14:07 |
| andreaf | sdague: I understand that's not something new - but in my experience at least it has not been so bad | 14:07 |
| *** yangyapeng has joined #openstack-qa | 14:08 | |
| andreaf | sdague: well we're working on this - jordanP already removed volume v1 tests and we'll be pruning scenarios / heavier sooner than later | 14:08 |
| sdague | andreaf: so, if you polled all 2000 developers on their sentiment on if the tempest gating jobs help them or hurt them in getting quality work done, what do think the results of said survey would be? | 14:09 |
| andreaf | sdague: I will discuss in the QA meeting tomorrow about stopping to get new tests in until we get this sorted, at least for resource consuming ones | 14:09 |
| *** yamahata has joined #openstack-qa | 14:09 | |
| andreaf | sdague: well we could do that poll | 14:09 |
| sdague | andreaf: yep, just be ready for the results | 14:10 |
| andreaf | sdague: but I'm also sure a number of developers would have no gating at all - so the result could be a bit biased | 14:10 |
| sdague | andreaf: sure, but the whole point is that unless the community as a whole sees this as possitive, we won't be able to address a bunch of issues | 14:11 |
| andreaf | heh the QA session for collecting input / feedback was an attempt of gathering this kind community feeling - I will sure do more of this type of exercise | 14:12 |
| sdague | and if the community sees the gate as the enemy (or parts of it), then they will just do whatever is required to get around it to get things in. Like rechecking realistically bad code because they don't trust a -1, and having some of it race in. | 14:12 |
| andreaf | sdague: I want to make sure the QA works on what the community feels as pain points | 14:13 |
| sdague | which actually happens a reasonable amount | 14:13 |
| *** nadya has joined #openstack-qa | 14:13 | |
| *** hferenc has joined #openstack-qa | 14:13 | |
| jordanP | yeah, so I share sdague views. We do a bad job right now, it's unacceptable. And people will start to do bold things if we don"t fix it now, like in real soon | 14:13 |
| sdague | andreaf: ok, so easy, stop failing good patches :) | 14:13 |
| sdague | jordanP: again, I think the tense is wrong :) | 14:13 |
| sdague | people already are | 14:13 |
| jordanP | yeah ok, maybe we need to get slapped harder :) | 14:14 |
| *** mpaolino has joined #openstack-qa | 14:14 | |
| jordanP | we do a bad job, yes. andreaf it's now impossible to have anything merged in tempest | 14:14 |
| sdague | a couple of years ago when I was spending a lot of time on the gate, I would watch the comment streams | 14:14 |
| andreaf | sdague: I'm sure you are annoyed by this with no tense :D | 14:14 |
| *** grafuls has joined #openstack-qa | 14:14 | |
| *** grafuls has quit IRC | 14:14 | |
| jordanP | so there's a big problem even for Tempest | 14:15 |
| jordanP | we need solution right now | 14:15 |
| sdague | jordanP: yeh | 14:15 |
| jordanP | what's annoying is our failures rates now is due to libvirt, memory issue and cirros weird thing (SSH) | 14:15 |
| sdague | so, I do really wonder how much removing the load on the scenario tests is going to help - https://review.openstack.org/#/c/438562/ | 14:15 |
| jordanP | sdague, the memory issue is related to Apache (for TLS) and the NOva api placement (yet another daemon) | 14:16 |
| jordanP | that's my claim, no proof, prove me wrong :) | 14:16 |
| jordanP | then libvirt, is buggy, what can the QA team do about it ? | 14:16 |
| sdague | jordanP: so, the OOM thing ... is going to be it's own beast | 14:16 |
| sdague | jordanP: run it less hard | 14:16 |
| jordanP | if SSH is not reliable, who's fault is it ? | 14:17 |
| andreaf | gate-tempest-dsvm-neutron-dvr-multinode-full-ubuntu-xenial-nv took 2h 27min on that patch?!? | 14:17 |
| jordanP | we have no leverage here | 14:17 |
| sdague | basically all the libvirt / qemu races we've reported before have gotten push back from upstream that they've never seen it | 14:17 |
| sdague | and production clouds don't either | 14:17 |
| jordanP | so we have to resort to workarounds, bad desperate moves (serialization, don't run scenarios at all etc...) | 14:17 |
| sdague | we're litterally running an order of magnitude hotter in gate than normal folks do | 14:18 |
| jordanP | which I am willing to do | 14:18 |
| sdague | andreaf: yeh, that's still running all 46 scenarios | 14:18 |
| *** mriedem_away is now known as mriedem | 14:18 | |
| sdague | I honestly think that to drop to that model you have to pick the 10 most important scenarios and not run the rest | 14:18 |
| andreaf | yeah but gate-tempest-dsvm-neutron-full-ubuntu-xenial is 1h 16min | 14:18 |
| sdague | andreaf: right, same deal | 14:19 |
| jordanP | but we need to stop the fire right now. So serial run is a good option. So is disabling all tests that perform SSH validation | 14:19 |
| andreaf | I guess the multinode runs migration tests or so? | 14:19 |
| sdague | andreaf: yeh, I don't know all the internals, it was something to post to see what it looks like | 14:19 |
| *** felipemonteiro_ has joined #openstack-qa | 14:20 | |
| *** felipemonteiro__ has joined #openstack-qa | 14:20 | |
| openstackrecheck | Console logs not available after 20:13s for gate-trove-python35-db 419248,3,c2ca412 | 14:21 |
| jordanP | andreaf, so ? what's our next move ? | 14:21 |
| andreaf | jordanP, sdague: we can make all scenario tests honour the run_ssh flag (which is false by default) | 14:21 |
| andreaf | which means things will break on ssh side even more :S | 14:22 |
| jordanP | that would be a simple skip_unless/skip_if decorator | 14:22 |
| andreaf | but at least it would calm things down and give us a bit more time | 14:22 |
| sdague | andreaf: so, anything that doesn't reduce load on the system during the scenario tests is I think kind of pointless | 14:22 |
| *** dave-mccowan has joined #openstack-qa | 14:22 | |
| jordanP | yeah ssh issues are due to very high load, that's my intuition too | 14:22 |
| sdague | the other thing we have to remember is that this cirrors guest running dropbear isn't like the most resilient timing system ever | 14:23 |
| sdague | so running that under a high load environment, isn't going to surprise me that when it gets scheduled out for a while during a heavy IOPs volume create, it f's up connectivity | 14:23 |
| sdague | you have to get the load under control first | 14:24 |
| jordanP | yeah, I did a 'openstack console log $cirros_img" locally (cirros running in a VM in my laptop) and compared with the logs outputed when there's a SSH failure in the gate. Cirros in the Gate ran super super slow, like 10 times slower | 14:25 |
| * andreaf brb | 14:25 | |
| jordanP | so maybe in dropbear there's some timing assumptions and we are just too slow | 14:25 |
| *** shausy has quit IRC | 14:25 | |
| *** amoralej|lunch is now known as amoralej | 14:25 | |
| *** jdandrea has joined #openstack-qa | 14:26 | |
| sdague | yeh, busybox is really meant for lightly loaded embedded environments where if things go wrong it just watchdog reboots and hopes for the best | 14:26 |
| sdague | and, when we think about coverage, if we can keep real sshing, but only 10 scenario tests, that's a better trade off in my head then keep 46 scenarios but turn of guest validation | 14:27 |
| jordanP | Right now https://review.openstack.org/#/c/438562/ I am in favor of | 14:27 |
| jordanP | if you can get 3 succesful Jenkins +1 | 14:28 |
| sdague | jordanP: yeh, I'd only be ok with that if we drop some scenarios | 14:28 |
| jordanP | which nowadays is impossible without that patch | 14:28 |
| andreaf | jordanP: yeah that takes too long as us | 14:28 |
| andreaf | as is | 14:28 |
| jordanP | not | 14:28 |
| sdague | because I do agree the run time is too long | 14:28 |
| jordanP | slow is better than false negative right ? | 14:28 |
| sdague | jordanP: yeh, I ran recheck again | 14:28 |
| jordanP | we need to drop some scenarios then | 14:28 |
| andreaf | jordanP, sdague: so what I'm doing is isolating which tests are consuming most resources | 14:28 |
| *** dimtruck is now known as zz_dimtruck | 14:29 | |
| andreaf | so we can skip scenarios in a smart way | 14:29 |
| jordanP | 2017-02-28 15:08:45.853270 | {0} tempest.scenario.test_network_v6.TestGettingAddress.test_dualnet_multi_prefix_dhcpv6_stateless [177.885508s] ... ok | 14:29 |
| jordanP | 2017-02-28 15:11:21.091682 | {0} tempest.scenario.test_network_v6.TestGettingAddress.test_dualnet_multi_prefix_slaac [155.209479s] . | 14:29 |
| *** zz_dimtruck is now known as dimtruck | 14:29 | |
| jordanP | 5min is way too much | 14:29 |
| sdague | jordanP: sure... but we are definitely going to be close to bumping up to 3 hours I think if we don't trim | 14:29 |
| sdague | andreaf: no, please start the other way around, rank the importance of the scenario tests 1 - N | 14:29 |
| *** ihrachys has joined #openstack-qa | 14:29 | |
| *** porrua has joined #openstack-qa | 14:30 | |
| jordanP | and who needs IPv6 and Encryption volume tests to run everytime everywhere ? | 14:30 |
| sdague | then we can just define a cutline at time | 14:30 |
| sdague | jordanP: ++ | 14:30 |
| sdague | or, more importantly, run 1 IPv6 test | 14:30 |
| sdague | so we know that ipv6 networking is a thing | 14:30 |
| sdague | but we don't need every allocation approach | 14:30 |
| jordanP | so in the end yeah, the way forward is to skip many scenarios and run scenarios serialy | 14:32 |
| sdague | andreaf: if you only start from trimming, vs. start from "these are the most important things" then we'll never actually have the conversation about why some things are less important, because we never enumerated what was the most important stuff | 14:33 |
| jordanP | I am +2 on that though I don"t like it | 14:33 |
| jordanP | because it seems Tempest has to pay the price for libvirt and services being too greedy | 14:33 |
| sdague | it is important that we know we can: boot from volume, have working ipv6, ssh into a created guest | 14:34 |
| sdague | jordanP: it's not really that | 14:34 |
| sdague | it's that testing is creating a model of the environment, and trying to do things to it | 14:34 |
| sdague | but our model currently is a system that's more busy (maybe even 10x more) than any real functioning system in the wild | 14:35 |
| sdague | so our model is skewed from reality | 14:35 |
| sdague | which means our asserts and validation aren't a good approximation for the real world | 14:36 |
| sdague | which is what we would like | 14:36 |
| jordanP | that current model has had real succes, like we found tons of race conditions and complicated scenarios that would expose a bug | 14:36 |
| sdague | jordanP: it has, and running some api tests in parallel is definitely important | 14:37 |
| jordanP | I don't think the model has to be changed, but we have lacked disciplines (Tempest and openstack services) so that led to side effects | 14:37 |
| sdague | but the scenario tests have rarely been the source of good bug resolution | 14:37 |
| sdague | most of that has come from more targetted things | 14:37 |
| sdague | .... and.... | 14:37 |
| sdague | projects now have a *lot* more testing of those kinds of things at a functional level, where they are easier to flush out | 14:38 |
| openstackgerrit | Ihar Hrachyshka proposed openstack-infra/devstack-gate master: Switch from lib/neutron-legacy to lib/neutron https://review.openstack.org/436798 | 14:38 |
| *** dimtruck is now known as zz_dimtruck | 14:38 | |
| andreaf | sdague: I fear I don't have a good picture about where functional testing is at in projects (at least in the 6 ones hosted in Tempest) | 14:39 |
| jordanP | nova and cinder are doing it, with success I believe | 14:39 |
| sdague | they all are | 14:40 |
| openstackgerrit | Felipe Monteiro proposed openstack/patrole master: Compute API Create Backup Test https://review.openstack.org/438676 | 14:40 |
| *** Swami has joined #openstack-qa | 14:40 | |
| *** Swami_ has joined #openstack-qa | 14:40 | |
| sdague | all have slightly different approaches, but they all are getting solid data that wasn't there when tempest switched to parallel | 14:40 |
| jordanP | andreaf, do you also agree that we should skip some scenario tests asap ? | 14:41 |
| openstackrecheck | Console logs not available after 20:06s for gate-functional-dsvm-magnum-swarm-ubuntu-xenial 436559,2,4424486 | 14:41 |
| andreaf | jordanP: yes I think we need to trim and get the system load under control | 14:41 |
| *** tonytan4ever has joined #openstack-qa | 14:41 | |
| andreaf | let | 14:42 |
| sdague | so I guess the question is would you rather have 10 scenario tests + ssh validation, or 46 without? I think that's kind of the concrete decision | 14:42 |
| jordanP | good, so I'll create an etherpad with all the tests then we will vote | 14:42 |
| andreaf | sdague: I'd rather keep ssh on | 14:42 |
| jordanP | I say 15 + ssh validation | 14:42 |
| jordanP | good | 14:42 |
| sdague | jordanP: sounds great, etherpad away | 14:42 |
| andreaf | jordanP: well I'm not sure about the exact number but let's start prioritising | 14:43 |
| andreaf | jordanP: we need to get the PTLs of the different projects to weigh in as well | 14:43 |
| sdague | jordanP: probably worth posting the timings from the serial run as well | 14:43 |
| jordanP | why ? (lol) | 14:43 |
| andreaf | about coverage in functional jobs and relevance | 14:43 |
| jordanP | yeah, that etherpad is going to include the timing from serial | 14:43 |
| andreaf | jordanP: ethercalc perhaps? | 14:44 |
| andreaf | so we can sort | 14:44 |
| jordanP | ok | 14:44 |
| sdague | andreaf: totally disagree that PTLs need to sign off on this one, I think this needs to be a "here is what we're doing on reliability front, this might reduce some coverage, but it will make things a bunch more stable. Lets discuss any coverage concerns now that this is done and figure out the best way to cover that (be it in tempest or in project functional tests)" | 14:44 |
| frickler | so I think I can spot an issue related to the SSH issues here: http://logs.openstack.org/60/439560/3/check/gate-tempest-dsvm-neutron-full-ubuntu-xenial/569fcd0/console.html#_2017-03-01_13_32_30_829083 | 14:45 |
| sdague | well, I think sort is less important than setting a time threshold | 14:45 |
| sdague | but sure | 14:45 |
| jordanP | where's the ethercalc thing ? | 14:45 |
| andreaf | sdague: I didn't mean sign off | 14:45 |
| sdague | https://ethercalc.openstack.org/ | 14:45 |
| jordanP | thanks | 14:45 |
| andreaf | sdague: I meant give an opinion on the relevance of tests and functional coverage | 14:45 |
| frickler | seems like there are a couple of connections attempts with failing auth, after that there is "connection reset by peer". I get the same behaviour if I have more then five parallel SSH sessions to a cirros instance | 14:46 |
| frickler | which might mean that paramiko doesn't close connections properly after an auth failure and tempest would need to add an ssh.close() | 14:46 |
| mtreinish | frickler: hmm, that's interesting. I think kevinbenton was looking at an issue with the metadata server around that | 14:46 |
| mtreinish | frickler: that should be pretty easy to add to tempest | 14:47 |
| felipemonteiro_ | Quick nooby question: What's the command for retriggering a test run for Citrix XenServer CI in a patchset? | 14:47 |
| andreaf | it's in the gerrit comment from citrix | 14:47 |
| frickler | mtreinish: somewhere around http://git.openstack.org/cgit/openstack/tempest/tree/tempest/lib/common/ssh.py#n114 I guess | 14:47 |
| frickler | mtreinish: could you give that a try for me? I'll be afk for a bit | 14:48 |
| *** salv-orlando has joined #openstack-qa | 14:48 | |
| andreaf | felipemonteiro_ : only when it fails though... | 14:48 |
| openstackgerrit | Denis Buliga proposed openstack-dev/devstack stable/ocata: Ensure we install setuptools from requirements https://review.openstack.org/439630 | 14:48 |
| mtreinish | frickler: sure, I can try to hack something together | 14:49 |
| andreaf | felipemonteiro_ : xenserver: recheck | 14:49 |
| jordanP | frickler, good finding, it's worth a try. Boy that common/ssh.py is a nested can of worms... | 14:49 |
| andreaf | frickler: yeah that sounds like a good thing to try | 14:50 |
| openstackgerrit | Denis Buliga proposed openstack-dev/devstack stable/newton: Ensure we install setuptools from requirements https://review.openstack.org/439631 | 14:50 |
| mtreinish | jordanP: it's paramiko in general. It's more headache than it's worth | 14:50 |
| sdague | well, it's also the way it's plumbed into the super classes for multiple uses | 14:50 |
| andreaf | mtreinish: do you have a chance to make a patch for that? | 14:51 |
| mtreinish | andreaf: I'm about to push one | 14:51 |
| andreaf | mtreinish: thanks | 14:51 |
| jordanP | mtreinish, please test it locally :D :D | 14:52 |
| mtreinish | jordanP: heh, that's why I haven't pushed it yet | 14:52 |
| *** rossella_s has quit IRC | 14:53 | |
| *** rossella_s has joined #openstack-qa | 14:54 | |
| *** mvk has joined #openstack-qa | 14:54 | |
| *** dustins has joined #openstack-qa | 14:57 | |
| *** nadya has quit IRC | 14:58 | |
| *** zz_dimtruck is now known as dimtruck | 14:59 | |
| jordanP | my ethercalc skills suck: https://ethercalc.openstack.org/nu56u2wrfb2b andreaf could you make it beautiful for me please ? | 14:59 |
| *** dustins_ has joined #openstack-qa | 15:01 | |
| andreaf | jordanP: lol it looks just fine | 15:01 |
| openstackrecheck | Console logs not available after 20:09s for gate-rally-dsvm-neutron-rally-ubuntu-xenial 415608,4,2b0be1e | 15:01 |
| jordanP | I am looking to have less precision in the timing measurements | 15:01 |
| andreaf | jordanP: my ethercal skills go little behind knowing about the existance of the tool | 15:01 |
| jordanP | 155.209479 doesn't make sense | 15:01 |
| andreaf | ok | 15:01 |
| openstackgerrit | Matthew Treinish proposed openstack/tempest master: Close connection after each failed connect attempt https://review.openstack.org/439638 | 15:02 |
| mtreinish | frickler, jordanP, andreaf: ^^^ | 15:02 |
| jordanP | andreaf, don"t be modest. We've seen what you are capable of at the PTG | 15:02 |
| mtreinish | I tested it locally by using the ssh module to try and login to one of my machines with the wrong password | 15:02 |
| *** dustins has quit IRC | 15:02 | |
| *** nadya has joined #openstack-qa | 15:03 | |
| *** mtanino has joined #openstack-qa | 15:03 | |
| jordanP | andreaf, you're good :) | 15:04 |
| sdague | jordanP: how do we get precision set in there? | 15:04 |
| sdague | I wanted a minutes column | 15:04 |
| jordanP | sdague, I didnt do it, andreaf did i think | 15:05 |
| andreaf | yeah set the format | 15:05 |
| *** awaugama has joined #openstack-qa | 15:05 | |
| jordanP | lol I clicked the format several times, never found how to do it.. | 15:05 |
| sdague | actually, there is a round function | 15:06 |
| sdague | andreaf: please stop adjusting precision :P | 15:06 |
| sdague | andreaf: can you undo whatever precision change you made on minutes | 15:06 |
| andreaf | sdague: ok | 15:07 |
| andreaf | sdague: it's back to default | 15:07 |
| sdague | andreaf: I would honestly also suggestion to do a 0.00 on the seconds column for format | 15:08 |
| sdague | it's not sticking for me | 15:08 |
| sdague | injecting string units into numbers is just super confusing from understanding how to do math with them | 15:08 |
| mtreinish | jordanP, andreaf: are you trying to make the tests table from: http://status.openstack.org/openstack-health/#/job/gate-tempest-dsvm-neutron-full-ssh ? | 15:10 |
| andreaf | sdague: whatever - I don't mind, I like units on the string | 15:10 |
| jordanP | mtreinish, no from sdague's patch here: https://review.openstack.org/#/c/438562/ | 15:11 |
| andreaf | mtreinish: not sure jordanP did | 15:11 |
| jordanP | not sure what job I took though... | 15:12 |
| jordanP | that may no be representativ | 15:12 |
| *** itzikb has quit IRC | 15:12 | |
| *** awaugama has quit IRC | 15:13 | |
| *** awaugama has joined #openstack-qa | 15:14 | |
| jordanP | real ethercalc gurus performing live, much magic happening :) | 15:15 |
| sdague | it's times like these where I wish this was a google drive doc | 15:16 |
| sdague | I actually know more about the function set you can use over there | 15:16 |
| jordanP | =IF(NOT(ISBLANK(E5)),C5,0) <--- that's not bad | 15:16 |
| openstackgerrit | Jeremy Zhang proposed openstack/tempest master: Small fix for test_volume_swap https://review.openstack.org/439648 | 15:16 |
| sdague | jordanP: yeh, though in google I could do it without the dedicated timing column | 15:16 |
| sdague | but I don't know how to manip ranges like that here | 15:16 |
| sdague | anyway, if we make double columns like that, we can figure out what choices we are making | 15:17 |
| andreaf | yeah I would prefer googledocs as well, but it's not accessible by everyone | 15:17 |
| sdague | or proposing | 15:17 |
| sdague | that would be my starting point for a minimum set | 15:19 |
| *** itzikb has joined #openstack-qa | 15:21 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Use modern PBR and hacking packages https://review.openstack.org/439649 | 15:21 |
| openstackrecheck | Console logs not available after 20:13s for gate-tempest-dsvm-networking-odl-boron-snapshot-v2driver 435813,14,989e88e | 15:21 |
| sdague | jordanP / andreaf ok, if you put any mark in your dedicated column, it will start adding up for you | 15:24 |
| jordanP | coool ! | 15:24 |
| jordanP | mtreinish, this is where the timings come from : http://logs.openstack.org/62/438562/1/check/gate-tempest-dsvm-neutron-dvr-multinode-full-ubuntu-xenial-nv/4a70259/console.html | 15:25 |
| sdague | anyone else can copy that number column, they just have to reset the counting column to C | 15:25 |
| sdague | I don't know how to keep that fixed in the formula language | 15:26 |
| mtreinish | jordanP: ok, I don't have any subunit2sql data I can compare to for that job, because it's non-voting and check only | 15:26 |
| andreaf | sdague: I guess $C does not work? | 15:26 |
| sdague | andreaf: I don't know | 15:26 |
| andreaf | sdague: it seems to | 15:27 |
| sdague | jordanP: also, while not rigorously statistically accurate, I noticed that the tests run noticably faster in serial, like drop 20s on a 2, 2.5 minute test vs. a comparable run I was looking at | 15:28 |
| sdague | which makes total sense with not competing over cpu iops | 15:28 |
| sdague | so the serial timings are a better baseline for what they would look like vs. anything we pulled from existing runs | 15:28 |
| *** mriedem1 has joined #openstack-qa | 15:29 | |
| sdague | I wish I knew how to stipple the spreadsheet, but all the ethercalc docs just seem to be "here is this awesome thing we made" and not really any user facing stuff | 15:30 |
| dims | sdague : can we get this in please? https://review.openstack.org/#/c/427392/ probably will help unblock ceilometer gate | 15:30 |
| dims | sileht : ^ | 15:31 |
| sileht | dims, I was asking the same :) | 15:31 |
| *** Swami_ has quit IRC | 15:31 | |
| andreaf | sdague: sorry what does stipple mean applied to a spreadsheet? | 15:31 |
| *** mriedem has quit IRC | 15:31 | |
| sdague | where the background rows are colored even/odd | 15:31 |
| *** Swami_ has joined #openstack-qa | 15:31 | |
| sdague | easier to read across horiz | 15:31 |
| sdague | dims / sileht let me look and make sure it's not just a failure shuffle | 15:32 |
| andreaf | hah | 15:32 |
| andreaf | thanks | 15:32 |
| *** artom_ is now known as artom | 15:32 | |
| sdague | mriedem1: sileht: is the drop of the || true on https://review.openstack.org/#/c/427392/1/lib/nova going to break cells v1? | 15:33 |
| *** Serlex has quit IRC | 15:33 | |
| *** Swami has quit IRC | 15:34 | |
| sileht | sdague, I don't think, nova hardcode 'return 0' | 15:35 |
| *** timello has quit IRC | 15:36 | |
| *** mpaolino has quit IRC | 15:36 | |
| sdague | sileht: no, I think it will throw another kind of exception | 15:37 |
| openstackgerrit | Merged openstack-infra/subunit2sql master: Use modern PBR and hacking packages https://review.openstack.org/439649 | 15:37 |
| *** timello has joined #openstack-qa | 15:37 | |
| *** mriedem1 is now known as mriedem | 15:38 | |
| mriedem | sdague: no map_cell0 is also needed for cells v1 | 15:38 |
| sileht | sdague, oh, I can't talk then | 15:38 |
| sdague | mriedem: ok | 15:38 |
| mriedem | and it's idempotent since ocata | 15:38 |
| sdague | mriedem: ok, that's fine, it wasn't super clear from the code | 15:38 |
| sdague | +2 | 15:38 |
| mriedem | sdague: https://github.com/openstack/nova/commit/aa7b6ebbb254f00fcb548832941ca9dbd3996d9f was the fix | 15:39 |
| sdague | mriedem: ok | 15:39 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add new graph time for variance of individual tests https://review.openstack.org/230611 | 15:39 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add migration to add common metadata fields to runs table https://review.openstack.org/334556 | 15:40 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add a new db manage command to migrate run_metadata https://review.openstack.org/334557 | 15:40 |
| sdague | mtreinish: you want to pull the trigger on that devstack change? | 15:41 |
| *** hongbin has joined #openstack-qa | 15:41 | |
| *** cody-somerville has joined #openstack-qa | 15:41 | |
| *** cody-somerville has quit IRC | 15:41 | |
| *** cody-somerville has joined #openstack-qa | 15:41 | |
| *** tonytan_brb has joined #openstack-qa | 15:41 | |
| openstackrecheck | Console logs not available after 20:10s for gate-tempest-dsvm-nnet-ubuntu-xenial 435224,6,151033d | 15:41 |
| *** tonytan4ever has quit IRC | 15:42 | |
| *** csomerville has joined #openstack-qa | 15:42 | |
| *** fnaval has joined #openstack-qa | 15:42 | |
| mtreinish | sdague: done | 15:42 |
| andreaf | sdague, jordanP: I think it's good to have a first approx about what we consider indispensable - but I we still need to check things like coverage in functional tests and failure rates before removing something from the integrated gate | 15:44 |
| jordanP | I disagree | 15:44 |
| jordanP | failure rates should be 0 so no question to ask here | 15:45 |
| sdague | andreaf: I moderately disagree, because I think that it's completely within the scope of the QA team to make this decision | 15:45 |
| andreaf | if a test does not have coverage anywhere else and it caught bugs before I would not want to remove it | 15:45 |
| jordanP | and coverage in func tests should be a concern to the project, not Tempest | 15:45 |
| sdague | the QA team adds tests all the time without consulting coverage in projects about duplication | 15:45 |
| jordanP | me neither, but am not responsible for the whole openstack galaxy | 15:45 |
| *** cody-somerville has quit IRC | 15:45 | |
| jordanP | we will inform with email and such that this and these are going to be skipped and projects will do what's necessary | 15:46 |
| sdague | jordanP: right, and I think that's the important bit, full coverage is expected to be a project level concern | 15:46 |
| sdague | and ... this being early in a release cycle, it's exactly the time to do this | 15:46 |
| andreaf | sdague: sure I agree it's under QA control but I think it would be good to check | 15:46 |
| sdague | so that people can reset to a new norm that tempest will run a bit lighter | 15:46 |
| sdague | if this was Milestone 3, I'd agree peeling back some tests is the wrong call, because you change the test parameters for the release | 15:47 |
| sdague | but now is the time to do it | 15:47 |
| andreaf | sdague: well right but also being early means more code is being written? so easier for regressions to get through? | 15:47 |
| *** yamahata has quit IRC | 15:47 | |
| sdague | andreaf: which is why teams get a heads up, and they get to look | 15:49 |
| andreaf | sdague, jordanP: do you guys want to get patches prepared about this and then we can discuss further on them and see how they behave in terms of time and stability | 15:49 |
| sdague | I think the difference of pov here is you assume that all these tests provide substantial coverage and want data proving they don't before doing things | 15:49 |
| sdague | and I assume they are pretty much not, as no data was ever really provided that this was the only/best way to test these things | 15:50 |
| sdague | we don't require new test submissions to demonstrate why that can't be tested in project trees | 15:50 |
| *** gamado has joined #openstack-qa | 15:50 | |
| andreaf | sdague: I don't assume they provide coverage, I simply don't know, and I'd like to get a better understanding before I accept to remove them | 15:50 |
| andreaf | sdague: because I'm not familiar with functional tests in tree | 15:51 |
| sdague | andreaf: right, and in the mean time, how much good code gets dropped on the floor because of the fail rate? | 15:52 |
| sdague | maybe I'd feel more comfortable if there was a deadline about how much longer the fail rate was going to be allowed before the big test cut | 15:53 |
| sdague | like Thursday EOD | 15:53 |
| *** DotanSoko has quit IRC | 15:53 | |
| sdague | but if this is like weeks of research, I think it's fundamentally the wrong call | 15:53 |
| jordanP | I'll be off for 10 days starting tomorrow | 15:53 |
| sdague | tempest is not supposed to be the first line of defence | 15:53 |
| sdague | it's supposed to be the last mile to just prove this all holds together as a whole | 15:54 |
| jordanP | andreaf, we don't want to remove tests, just skip them. | 15:54 |
| *** armax has quit IRC | 15:54 | |
| sdague | https://twitter.com/noahsussman/status/836612175707930625 | 15:54 |
| jordanP | nothing is set in stone | 15:54 |
| andreaf | jordanP: heh sure | 15:54 |
| sdague | and honestly, it's going to be a ton easier to cut deep here, let everyone know we're running light to try to get consistency, and drive the conversation of what has to be added back in | 15:55 |
| andreaf | sdague: sure that inverted cone makes absolute sense | 15:56 |
| sdague | because ... like the pg conversation, every attempt to remove things 1 at a time pretty much gets an argument about why you can't do that :) | 15:56 |
| *** gszasz has quit IRC | 15:56 | |
| *** gvrangan has joined #openstack-qa | 15:58 | |
| *** tonytan_brb has quit IRC | 15:59 | |
| *** tonytan4ever has joined #openstack-qa | 15:59 | |
| sdague | jordanP: so if you had to get back to < 15 minutes, what would you drop on your list | 16:00 |
| jordanP | sdague, I am on it | 16:00 |
| jordanP | I am not done yet | 16:00 |
| *** yamamoto has joined #openstack-qa | 16:00 | |
| jordanP | almost | 16:01 |
| sdague | ok | 16:01 |
| openstackrecheck | Console logs not available after 20:14s for gate-rally-dsvm-rally-ubuntu-xenial 437829,2,a4daab0 | 16:02 |
| frickler | mtreinish: I didn't think the patch could be that easy, I assumed one would be needing another try-wrapper. but I found more evidence that my suspicion will be correct, see https://bugs.launchpad.net/nova/+bug/1668958/comments/4 | 16:03 |
| openstack | Launchpad bug 1668958 in OpenStack Compute (nova) "metadata service occasionally not returning keys" [Undecided,New] | 16:03 |
| *** rcernin has quit IRC | 16:03 | |
| jordanP | test_minimum_basic_scenario is pretty much already tested by every other scenarios | 16:03 |
| sdague | jordanP: while I know the swift tests are super cheap, I honestly think they are so much better covered in their functional job that we probably shouldn't bother. But that's mostly philisophical, as it wouldn't really impact timings | 16:04 |
| jordanP | I am not sure about the anonymous download | 16:04 |
| sdague | also, just fyi, swift did run tempest in the gate at all from Oct -> last week due to a project config bug | 16:05 |
| sdague | and no one noticed :) | 16:05 |
| sdague | and nothing was broken | 16:06 |
| andreaf | sdague: didn't? | 16:06 |
| sdague | nope | 16:06 |
| *** yamamoto has quit IRC | 16:06 | |
| sdague | when the neutron default job naming bits were flipped around | 16:06 |
| mtreinish | frickler: yeah, I guess the close() can raise it's own exception which would break out of the loop. But I think that might be desirable, because something went horribly wrong in that case | 16:06 |
| sdague | it turns out swift only had the nnet jobs on it | 16:06 |
| sdague | which were turned off for ocata/pike | 16:06 |
| sdague | except in cellsv1 case | 16:07 |
| jordanP | it did or it didn't run Tempest ? | 16:07 |
| andreaf | sdague: well I'm not sure neutron or nnet makes much difference for swift? | 16:07 |
| sdague | jordanP: it did not | 16:07 |
| sdague | on ocata / pike | 16:07 |
| mtreinish | frickler: I guess we still need to figure out why the pubkey isn't getting set in the guest | 16:07 |
| *** derekjhyang has joined #openstack-qa | 16:07 | |
| sdague | for 4 months | 16:07 |
| mtreinish | frickler: but at least this way we won't have the banner errors anymore | 16:07 |
| jordanP | yeah, so what ? Swift folk don't care that much about Tempest, no surprise here :) | 16:07 |
| sdague | jordanP: right, but when we talk about "it's important to run this test for coverage" | 16:08 |
| sdague | and... we didn't run that test for 4 months on the project that could have broken it | 16:08 |
| * mtreinish probably just jinxed it | 16:08 | |
| sdague | and there was no issue | 16:08 |
| sdague | we need to check that first assumption | 16:08 |
| jordanP | yeah, because they didn't break stuff for 4 months, good for them | 16:08 |
| jordanP | had they had broken stuff, we would have noticed | 16:09 |
| andreaf | mtreinish: does it makes to push multiple copies of https://review.openstack.org/#/c/439638/ in parallel to get more testing done on it quicker? | 16:10 |
| *** gamado has quit IRC | 16:10 | |
| frickler | mtreinish: I think it is being set, but not fast enough. If I interpret the retry pattern correctly, the time frame is 23 seconds until the fifth failed auth request blocks everything that follows. in the example I posted, the uptime of the cirros instance is 25s when it has finished booting. | 16:10 |
| sdague | jordanP: sure, but if we can run for that long without the test and other things are preventing that regression, why would it be a pre-commit test on every test? | 16:10 |
| sdague | on every patch | 16:11 |
| mtreinish | andreaf: nah, I think once the gate gets back we should be good to move forward on it | 16:11 |
| mtreinish | frickler: haha, now your comment about it being so easy makes more sense :) | 16:12 |
| mtreinish | andreaf: although I probably should have put the bug number in the commit message | 16:12 |
| *** sshwarts has quit IRC | 16:12 | |
| jordanP | humm that close() thing makes me ashame lol | 16:14 |
| openstackgerrit | Merged openstack/patrole master: Compute API Server Password Test https://review.openstack.org/439139 | 16:14 |
| jordanP | I am done with my tests selection btw | 16:18 |
| jordanP | we have some disagreement but maybe I first step will be to skip tests that no one "selected" | 16:18 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/subunit2sql master: Add a new db manage command to migrate run_metadata https://review.openstack.org/334557 | 16:18 |
| jordanP | that should buy already ~20min | 16:19 |
| *** Serlex has joined #openstack-qa | 16:21 | |
| openstackrecheck | Console logs not available after 20:09s for gate-functional-dsvm-magnum-k8s-ubuntu-xenial 414838,13,818a65f | 16:22 |
| sdague | jordanP: why did you choose the other 2 bfv scenarios? | 16:23 |
| jordanP | bfv ? | 16:23 |
| jordanP | boot from volume | 16:23 |
| sdague | yeh | 16:24 |
| sdague | the ebs and snapshot | 16:24 |
| openstackgerrit | Matthew Treinish proposed openstack-dev/devstack master: Differentiate between DEFAULT_IMAGE_NAME and filename https://review.openstack.org/435106 | 16:24 |
| openstackgerrit | Matthew Treinish proposed openstack-dev/devstack master: clean up virt drivers that we do image logic for https://review.openstack.org/435126 | 16:24 |
| jordanP | test_volume_boot_pattern is already tested by the other 2 | 16:24 |
| jordanP | and booting from a volume snapshot must be tested | 16:24 |
| jordanP | because it creates a volume under the hood and boot from that volume | 16:25 |
| jordanP | which covers more than just "boot from a volume that a user precreated" | 16:25 |
| *** gamado has joined #openstack-qa | 16:26 | |
| jordanP | than the ebs thing because it's a real interesting test were a glance image actually references a snapshot but doesn"t actually holds the data | 16:27 |
| mtreinish | johnsom: on https://review.openstack.org/435106 did you have a link to the patch you were using to test octavia with? | 16:29 |
| sdague | jordanP: https://review.openstack.org/#/c/438562/ 3 clean rechecks :) | 16:30 |
| sdague | jordanP: so, honestly, I'd be fine taking your list | 16:30 |
| jordanP | that's good to hear, what do you think andreaf, mtreinish ? | 16:31 |
| jordanP | https://review.openstack.org/#/c/438562/ great, we need it then | 16:32 |
| *** mriedem has quit IRC | 16:34 | |
| sdague | jordanP: well the way it really should be done is as a tempest change that changes what -e full means | 16:34 |
| jordanP | yeah, that change shouldn"t happen in devstack-gate, indeed | 16:35 |
| *** bkopilov has quit IRC | 16:35 | |
| johnsom | mtreinish https://review.openstack.org/#/c/434518/ | 16:36 |
| *** gcheresh_ has quit IRC | 16:37 | |
| johnsom | mtreinish We ended up merging a hard coded DEFAULT_IMAGE_NAME to get our gates functional again: https://review.openstack.org/#/c/438680 | 16:37 |
| openstackgerrit | Sean Dague proposed openstack/tempest master: make -e full run tempest scenarios serially https://review.openstack.org/439698 | 16:37 |
| sdague | jordanP: so I think ^^^ is where we would add the skips to | 16:37 |
| *** martinkopec has quit IRC | 16:38 | |
| *** itzikb has quit IRC | 16:39 | |
| *** gamado has quit IRC | 16:41 | |
| openstackrecheck | Console logs not available after 20:16s for gate-swift-pep8-ubuntu-xenial 423906,4,4a4f8cf | 16:42 |
| *** salv-orlando has quit IRC | 16:44 | |
| *** mriedem has joined #openstack-qa | 16:44 | |
| *** cdent has joined #openstack-qa | 16:44 | |
| openstackgerrit | Merged openstack/patrole master: Changes tox to only run unit tests and moves unit tests to tests/unit. https://review.openstack.org/438324 | 16:46 |
| *** vhosakot has joined #openstack-qa | 16:46 | |
| hogepodge | mtreinish: this seems weird. http://paste.openstack.org/show/600970/ | 16:50 |
| *** markus_z has quit IRC | 16:50 | |
| *** xinli has joined #openstack-qa | 16:50 | |
| xinli | mtreinish: one question, from last week, a lot of cinder CIs broken, one of the issue is that devStack can't start on Ubuntu trusty, is there any scenario to fix it if devStack doesn't support it on trusty? | 16:52 |
| *** rcernin has joined #openstack-qa | 16:52 | |
| jordanP | xinli, 3rd party CI's must switch to use Ubuntu 16/Xenial | 16:53 |
| sdague | xinli: pike is not supported on trusty, and we've gotten to the point where we've merged features into projects that need newer libraries | 16:53 |
| xinli | jordanP: sdague: thanks | 16:54 |
| *** Swami_ has quit IRC | 16:56 | |
| *** salv-orlando has joined #openstack-qa | 17:00 | |
| *** tonytan4ever has quit IRC | 17:00 | |
| openstackgerrit | Nicolas Helgeson proposed openstack/tempest master: Identity V3-ext Oauth1 Consumers Client https://review.openstack.org/433818 | 17:00 |
| *** tonytan4ever has joined #openstack-qa | 17:01 | |
| *** asselin_ has quit IRC | 17:01 | |
| openstackrecheck | Console logs not available after 20:07s for gate-tempest-dsvm-neutron-full-ssh 437823,1,ac8b8a1 | 17:02 |
| *** jose-phillips has joined #openstack-qa | 17:02 | |
| *** yamamoto has joined #openstack-qa | 17:03 | |
| openstackgerrit | Felipe Monteiro proposed openstack/patrole master: Compute API Suspend Server Test https://review.openstack.org/439717 | 17:03 |
| *** nadya has quit IRC | 17:08 | |
| *** yamamoto has quit IRC | 17:08 | |
| *** nadya has joined #openstack-qa | 17:08 | |
| *** nadya has quit IRC | 17:09 | |
| dansmith | mtreinish: mriedem: remind me how to run one tempest test? this isn't working: tox -eall -- --regex test_server_basic_opts | 17:10 |
| *** breitz has joined #openstack-qa | 17:10 | |
| *** Apoorva has joined #openstack-qa | 17:11 | |
| mriedem | i don't think you want the -eall | 17:11 |
| mriedem | let me check the cells v1 job | 17:11 |
| *** e0ne has quit IRC | 17:11 | |
| mtreinish | hogepodge: that looks like a misconfigured cinder. There are 2 options you have to set when using ssl to make it return the right thing for links | 17:12 |
| mtreinish | hogepodge: clarkb had the same issue when we turned on ssl in the gate | 17:12 |
| mriedem | tempest run --regex ? | 17:12 |
| *** Swami_ has joined #openstack-qa | 17:12 | |
| mriedem | dansmith: http://logs.openstack.org/63/435563/2/check/gate-tempest-dsvm-cells-ubuntu-xenial/b144714/console.html.gz#_2017-02-17_20_25_44_673750 | 17:12 |
| dansmith | gdi I remember you RTFLog'ing me last time.. sorry | 17:12 |
| mtreinish | dansmith: drop the '--regex' | 17:12 |
| mtreinish | tox -eall -- test_server_basic_opts should work | 17:13 |
| *** aarefiev is now known as aarefiev_afk | 17:13 | |
| hogepodge | mtreinish: ah I see now. I missed the 's' | 17:13 |
| mriedem | mtreinish: would something with tox -e ostestr also do it? | 17:13 |
| *** agurenko has quit IRC | 17:13 | |
| mriedem | hmm https://github.com/openstack/tempest/blob/master/tox.ini#L48 | 17:14 |
| mriedem | tempest run --regex {posargs} | 17:14 |
| mriedem | yeah | 17:14 |
| mriedem | so just pass the thing in | 17:14 |
| mtreinish | mriedem: yeah you could also do tox -e ostestr -- --regex test_server_basic_opts and that should work too | 17:14 |
| dansmith | doesn't work | 17:14 |
| dansmith | http://pastebin.com/hkryE52D | 17:14 |
| clarkb | hogepodge: mtreinish ya that issue was :( there really should only be one thing but there is more | 17:14 |
| openstackgerrit | Nicolas Helgeson proposed openstack/patrole master: Identity V3 Tests - Roles https://review.openstack.org/437039 | 17:14 |
| mtreinish | dansmith: there is a stray '-' in the regex | 17:14 |
| *** felipemonteiro__ has quit IRC | 17:15 | |
| *** felipemonteiro_ has quit IRC | 17:15 | |
| mriedem | yeah -test_server_basic_opts | 17:15 |
| mriedem | drop that - | 17:15 |
| dansmith | still no, but maybe the regex just isn't matching | 17:15 |
| jordanP | it's basic "ops" | 17:15 |
| jordanP | not opts | 17:15 |
| clarkb | public_endpoint and osapi_volume_base_URL looks like | 17:15 |
| mriedem | dansmith: yeah wrong test name | 17:16 |
| mriedem | test_server_basic_ops | 17:16 |
| openstackgerrit | Felipe Monteiro proposed openstack/patrole master: Add role validation to Patrole framework when switching roles. https://review.openstack.org/438203 | 17:16 |
| dansmith | gdi | 17:16 |
| mriedem | http://git.openstack.org/cgit/openstack/tempest/tree/tempest/scenario/test_server_basic_ops.py#n129 | 17:16 |
| dansmith | why do I suck? | 17:16 |
| mriedem | maybe you need bifocals? | 17:16 |
| castulo | :P | 17:17 |
| openstackgerrit | Felipe Monteiro proposed openstack/patrole master: Required modification in Framework https://review.openstack.org/433741 | 17:17 |
| dansmith | I need fewer parallel tasks, is what I need | 17:18 |
| mtreinish | mriedem: if we want to get really fancy you could also do 'tox -eostestr -- --no-discover tempest.scenario.test_server_basic_ops' | 17:19 |
| *** felipemonteiro_ has joined #openstack-qa | 17:20 | |
| *** felipemonteiro__ has joined #openstack-qa | 17:20 | |
| mriedem | mtreinish: i drink with my pinky down thank you very much | 17:20 |
| mriedem | you can keep your new england elitist regex to yourself | 17:21 |
| mriedem | is it new england or just ivy league? | 17:21 |
| mriedem | i forget the insult | 17:21 |
| mtreinish | haha | 17:21 |
| openstackrecheck | Console logs not available after 20:27s for gate-fuxi-dsvm-fullstack 420386,4,27e6f4a | 17:23 |
| mtreinish | mriedem: do you have any thoughts on: https://review.openstack.org/#/c/439428/ | 17:30 |
| mriedem | "If we filter by 10.1.1.1 I feel like nova is misbehaving if it returns the server with: 10.1.1.10" | 17:31 |
| mtreinish | johnsom: ok, that's what I was going to recommend you do. Since it's better to just take a hard stance on that | 17:31 |
| mriedem | idk, it's regex | 17:31 |
| mtreinish | johnsom: I just think the octavia case is a good way for me to test, since you have an image which begins with 'a' | 17:32 |
| johnsom | mtreinish Ha, sad for us, it does start with 'a'.... | 17:32 |
| johnsom | mtreinish Keep me posted if we can pull out that hard coded cirros version. | 17:33 |
| mtreinish | mriedem: sure, I get that. It just feels weird | 17:34 |
| mtreinish | johnsom: well I'll likely push up a test patch at some point to verify that I'm doing things correctly | 17:34 |
| mriedem | mtreinish: hmm, looking at the test, i figured it was doing some kind of wildcarding | 17:35 |
| mriedem | but it's not | 17:35 |
| mriedem | i'd have to trace this down into the nova db api code | 17:35 |
| *** lucasagomes is now known as lucas-afk | 17:35 | |
| mriedem | if we're not asked to wildcard, but we're doing a LIKE or % or something in the DB API, then i could see maybe raising an error? or doing a re-query w/o the LIKE if there are multiple results | 17:36 |
| *** dtantsur is now known as dtantsur|afk | 17:39 | |
| openstackrecheck | Console logs not available after 20:14s for gate-functional-dsvm-magnum-k8s-ubuntu-xenial 438321,2,490e8b4 | 17:43 |
| sdague | I thought the layer4 test was made non voting? | 17:45 |
| sdague | because sahara plugin looks broken still | 17:45 |
| *** rfolco has joined #openstack-qa | 17:48 | |
| *** armax has joined #openstack-qa | 17:49 | |
| andreaf | mtreinish: all nice and green: https://review.openstack.org/#/c/439638/1 | 17:49 |
| mtreinish | andreaf: yeah | 17:50 |
| mtreinish | andreaf: I think we should land that change | 17:50 |
| andreaf | yeah ok | 17:50 |
| andreaf | +A | 17:50 |
| sdague | oh, it was a fatal git clone which didn't die early enough | 17:51 |
| andreaf | sdague: so layer4 only runs heat tests which are still in Tempest tree | 17:53 |
| sdague | andreaf: is there a reason it still enables the sahara plugin ? | 17:53 |
| *** salv-orlando has quit IRC | 17:54 | |
| sdague | and is there an eta on that ending? | 17:54 |
| andreaf | sdague: don't know I kind of lost track of the story of layer4 - the only good thing it does as far as I know is to test pre-provisioned credentials, which I would be happy to move to another job | 17:55 |
| andreaf | mtreinish: do you know about layer for and sahara? | 17:55 |
| * tosky hears sahara | 17:56 | |
| andreaf | tosky: heh perhaps you know about the layer4 job in Tempest and sahara? | 17:57 |
| *** salv-orlando has joined #openstack-qa | 17:57 | |
| sdague | andreaf: on the upside - https://review.openstack.org/#/c/439698/ seems to do the same thing as the d-g patch | 17:57 |
| tosky | andreaf: isn't layer4 the job which tests integration with some less core components? | 17:57 |
| sdague | got to run away for it bit, but andreaf I would propose taking jordanP's list and putting it into that patch | 17:58 |
| andreaf | sdague: ok I can follow up on that | 17:58 |
| jordanP | yeah, sorry I can"t work on this anymore, I have some issues at work here | 17:58 |
| sdague | jordanP: yep, no worries | 17:58 |
| andreaf | jordanP: no worries | 17:59 |
| sdague | jordanP: you built the list, I think that's good enough | 17:59 |
| tosky | andreaf: layer4 is not the place for sahara jobs now: https://review.openstack.org/#/c/328239/ | 17:59 |
| *** Swami_ has quit IRC | 17:59 | |
| andreaf | sdague, jordanP: I will be off the best part of the day tomorrow as well, so I will make something later tonight for that patch | 17:59 |
| tosky | andreaf: so if I understand it correctly, if heat is moved out from tempest, layer4 can be kiled | 17:59 |
| andreaf | ok that was my understanding as well | 18:00 |
| andreaf | tosky: thanks | 18:00 |
| tosky | talking about this, what is the status of the heat tests/tempest client? | 18:00 |
| tosky | I mean, is their move to heat/heat-tests/whatever!tempest close enough? | 18:00 |
| tosky | (if you discussed about this at the PTG) | 18:01 |
| andreaf | no unfortunately I didn't - oomichi or mtreinish may know more - but I'll take a note to follow up on this | 18:01 |
| *** tonytan4ever has quit IRC | 18:01 | |
| tosky | thanks! | 18:02 |
| *** tonytan4ever has joined #openstack-qa | 18:02 | |
| openstackrecheck | Console logs not available after 20:17s for gate-tempest-dsvm-neutron-full-ubuntu-xenial 437491,1,a63c450 | 18:03 |
| *** yamamoto has joined #openstack-qa | 18:04 | |
| *** jordanP has quit IRC | 18:06 | |
| *** armax has quit IRC | 18:06 | |
| openstackgerrit | Merged openstack-dev/devstack master: nova: call map_cell0 much earlier in the setup https://review.openstack.org/427392 | 18:08 |
| *** nadya has joined #openstack-qa | 18:09 | |
| *** yamamoto has quit IRC | 18:10 | |
| *** abehl has quit IRC | 18:10 | |
| openstackgerrit | Merged openstack-infra/devstack-gate master: Support setting a variable to get local.conf merged late https://review.openstack.org/438957 | 18:11 |
| *** dsariel has quit IRC | 18:12 | |
| openstackgerrit | Dan Smith proposed openstack-dev/devstack master: Move rpc backend start/configure to earlier https://review.openstack.org/437498 | 18:13 |
| openstackgerrit | Dan Smith proposed openstack-dev/devstack master: WIP: Fleetify nova conductor for N cells https://review.openstack.org/436094 | 18:13 |
| *** derekjhyang has quit IRC | 18:13 | |
| *** tosky has quit IRC | 18:13 | |
| *** nadya has quit IRC | 18:15 | |
| *** tesseract has quit IRC | 18:16 | |
| *** Swami has joined #openstack-qa | 18:18 | |
| oomichi | tosky: andreaf: I think heat tests should be moved into heat repo, and the clients and tessts should be removed from tempest. But the migration doesn't happen yet on heat side | 18:21 |
| openstackrecheck | Console logs not available after 20:11s for gate-swift-pep8-ubuntu-xenial 438322,1,5c07909 | 18:23 |
| *** asselin has joined #openstack-qa | 18:24 | |
| *** cdent has quit IRC | 18:29 | |
| *** Alex_Staf has quit IRC | 18:34 | |
| openstackgerrit | Jaewoo Park proposed openstack/patrole master: Users RBAC test for Keystone API v2 users https://review.openstack.org/439137 | 18:39 |
| *** e0ne has joined #openstack-qa | 18:40 | |
| *** jdandrea has quit IRC | 18:40 | |
| *** jdandrea has joined #openstack-qa | 18:41 | |
| *** harlowja has quit IRC | 18:43 | |
| openstackrecheck | Console logs not available after 20:09s for gate-tempest-dsvm-neutron-full-ssh 436842,1,ba32f4d | 18:44 |
| *** gszasz has joined #openstack-qa | 18:44 | |
| *** harlowja has joined #openstack-qa | 18:46 | |
| *** yamahata has joined #openstack-qa | 18:47 | |
| *** nadya has joined #openstack-qa | 18:48 | |
| sdague | andreaf: I can probably whip up the patch | 18:51 |
| andreaf | sdague: sure go ahead - I won't be able to do that until later tonight | 18:52 |
| sdague | andreaf: so, on the good side, the only 2 job failures in the list so far were - infra git clone failing | 18:52 |
| *** jdandrea has quit IRC | 18:52 | |
| sdague | and job timeout on a rax node that is super duper slow (for dvr job) | 18:52 |
| sdague | the dvr job also seems to only have a 120 backstop | 18:52 |
| *** salv-orl_ has joined #openstack-qa | 18:52 | |
| openstackgerrit | Emily Hugenbruch proposed openstack/tempest master: Change rebuild tests to use image_ref https://review.openstack.org/433162 | 18:53 |
| *** cdent has joined #openstack-qa | 18:54 | |
| *** gszasz has quit IRC | 18:54 | |
| *** salv-orlando has quit IRC | 18:55 | |
| *** gouthamr has quit IRC | 18:56 | |
| *** gouthamr has joined #openstack-qa | 19:00 | |
| *** tonytan4ever has quit IRC | 19:02 | |
| *** tonytan4ever has joined #openstack-qa | 19:03 | |
| openstackrecheck | Console logs not available after 20:12s for gate-tempest-dsvm-networking-midonet-v2-ubuntu-xenial 438075,3,f88f2fd | 19:04 |
| *** dharinic is now known as dharinic|lunch | 19:05 | |
| *** yamamoto has joined #openstack-qa | 19:06 | |
| *** shardy has quit IRC | 19:09 | |
| *** jdandrea has joined #openstack-qa | 19:10 | |
| *** yamamoto has quit IRC | 19:12 | |
| *** gfidente is now known as gfidente|afk | 19:17 | |
| *** pkoniszewski has quit IRC | 19:18 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Deprecate resources_prefix and change rand_name() https://review.openstack.org/439203 | 19:18 |
| openstackgerrit | Ihar Hrachyshka proposed openstack-dev/devstack master: Use lib/keystone to configure [nova] section in neutron.conf https://review.openstack.org/439787 | 19:19 |
| *** e0ne has quit IRC | 19:20 | |
| *** salv-orl_ has quit IRC | 19:23 | |
| openstackgerrit | Ihar Hrachyshka proposed openstack-infra/devstack-gate master: Switch from lib/neutron-legacy to lib/neutron https://review.openstack.org/436798 | 19:23 |
| *** pkoniszewski has joined #openstack-qa | 19:24 | |
| openstackrecheck | Console logs not available after 20:24s for gate-tempest-dsvm-neutron-full-ubuntu-xenial 438171,5,482fc20 | 19:24 |
| openstackgerrit | Ihar Hrachyshka proposed openstack-infra/devstack-gate master: Switch from lib/neutron-legacy to lib/neutron https://review.openstack.org/436798 | 19:25 |
| openstackgerrit | Ihar Hrachyshka proposed openstack-dev/devstack master: lib/neutron: set variables needed for sane l3 agent setup https://review.openstack.org/438672 | 19:28 |
| openstackgerrit | Ihar Hrachyshka proposed openstack-dev/devstack master: lib/neutron: untangle metering configuration from legacy https://review.openstack.org/438687 | 19:28 |
| openstackgerrit | Sean Dague proposed openstack/tempest master: make -e full run tempest scenarios serially https://review.openstack.org/439698 | 19:28 |
| sdague | andreaf: I think that's jordanP's list. There were 2 issues where smoke test tags ended up not on his list. I think in one case I left the test, and the other I moved the smoke tag to a related test. | 19:29 |
| sdague | That comes to 18m run time in jordan's list | 19:29 |
| *** stream10 has joined #openstack-qa | 19:35 | |
| *** jose-phillips has quit IRC | 19:40 | |
| *** Alex_Staf has joined #openstack-qa | 19:44 | |
| openstackrecheck | Console logs not available after 20:06s for gate-functional-dsvm-magnum-swarm-ubuntu-xenial 436559,2,9bca6de | 19:44 |
| *** xinli has quit IRC | 19:45 | |
| openstackgerrit | Merged openstack/tempest master: Remove unused client in setup_clients https://review.openstack.org/439473 | 19:46 |
| *** cdent has quit IRC | 19:46 | |
| *** tonytan4ever has quit IRC | 19:49 | |
| *** raildo has joined #openstack-qa | 19:49 | |
| *** tonytan4ever has joined #openstack-qa | 19:49 | |
| *** nadya has quit IRC | 19:53 | |
| *** Alex_Staf has quit IRC | 19:53 | |
| *** pcaruana has quit IRC | 19:58 | |
| *** armax has joined #openstack-qa | 19:59 | |
| *** dharinic|lunch is now known as dharinic | 20:02 | |
| *** openstackgerrit has quit IRC | 20:03 | |
| *** amoralej is now known as amoralej|off | 20:03 | |
| *** gcheresh_ has joined #openstack-qa | 20:04 | |
| openstackrecheck | Console logs not available after 20:11s for gate-tacker-pep8-ubuntu-xenial 438336,1,8906dbc | 20:05 |
| *** harlowja has quit IRC | 20:08 | |
| *** yamamoto has joined #openstack-qa | 20:08 | |
| *** pkoniszewski has quit IRC | 20:09 | |
| *** openstackgerrit has joined #openstack-qa | 20:10 | |
| openstackgerrit | Chi Lo proposed openstack/patrole master: Add Subnetpool test cases for RBAC. https://review.openstack.org/439827 | 20:10 |
| *** yamamoto has quit IRC | 20:13 | |
| openstackgerrit | Merged openstack/patrole master: Compute API Suspend Server Test https://review.openstack.org/439717 | 20:14 |
| *** pkoniszewski has joined #openstack-qa | 20:17 | |
| *** jose-phillips has joined #openstack-qa | 20:19 | |
| *** xinli has joined #openstack-qa | 20:19 | |
| openstackgerrit | Felipe Monteiro proposed openstack/patrole master: Required modification in Framework https://review.openstack.org/433741 | 20:23 |
| openstackgerrit | Ihar Hrachyshka proposed openstack-dev/devstack master: Configure bridge_mappings for ovs and linuxbridge agents https://review.openstack.org/439829 | 20:24 |
| openstackrecheck | Console logs not available after 20:07s for gate-dib-dsvm-functests-ubuntu-xenial 438334,1,6e72a6d | 20:25 |
| openstackgerrit | Ihar Hrachyshka proposed openstack-infra/devstack-gate master: Switch from lib/neutron-legacy to lib/neutron https://review.openstack.org/436798 | 20:25 |
| openstackgerrit | Matthew Treinish proposed openstack/tempest master: WIP: Add tempest test writing guide https://review.openstack.org/439830 | 20:28 |
| mtreinish | mriedem: ^^^ it's still incomplete, but that's something I think you can help me with | 20:28 |
| mriedem | i can help -1 it | 20:29 |
| mtreinish | hah | 20:29 |
| mtreinish | I was thinking more about if the content about the setUpClass magic was clear enough | 20:30 |
| mtreinish | and where I need to add more details | 20:30 |
| *** krtaylor has quit IRC | 20:32 | |
| *** DotanSoko has joined #openstack-qa | 20:34 | |
| mtreinish | frickler: so the banner error is fixed, but it looks like there is still a situation where the pubkey isn't showing up in the guest: http://logs.openstack.org/38/439638/1/gate/gate-tempest-dsvm-neutron-full-ubuntu-xenial/2cf61a4/console.html#_2017-03-01_20_31_59_063287 | 20:37 |
| mtreinish | kevinbenton: ^^^ | 20:37 |
| mtreinish | that's on the gate run for the patch adding the close() | 20:37 |
| EmilienM | can someone remind me why we removed Javelin and where can I use a similar tool now? | 20:39 |
| EmilienM | a tl;dr works :) | 20:39 |
| clarkb | mtreinish: tempest uses paramiko right? the close() reminded me that infra had to explicitly close its paramiko connections or they leak | 20:40 |
| mtreinish | EmilienM: it was removed because it was basically the same thing as the openstack modules in ansible (just not as mature) and grenade just uses some bash to call osc to do the same thing now | 20:41 |
| clarkb | mtreinish: this iwth new releases from paramiko | 20:41 |
| mtreinish | clarkb: yeah, tempest is using paramiko | 20:41 |
| clarkb | mtreinish: you likely don't open enough threads to really notice but probably not a bad idea to explictily close used connections | 20:41 |
| mtreinish | clarkb: in this case we were leaving dangling connections on connection errors and drop bear falls over with > 5 open connections (hence the banner errors) | 20:41 |
| mtreinish | clarkb: https://review.openstack.org/439638 | 20:42 |
| EmilienM | mtreinish: if I was a Javelin user, what can I do now? | 20:42 |
| clarkb | mtreinish: nice, in our case we ended up with thousands of threads all fighting for cpu | 20:42 |
| mtreinish | clarkb: https://bugs.launchpad.net/nova/+bug/1668958/comments/4 | 20:42 |
| openstack | Launchpad bug 1668958 in OpenStack Compute (nova) "metadata service occasionally not returning keys" [Undecided,New] | 20:42 |
| EmilienM | mtreinish: is there a project/module that does the same hting? | 20:42 |
| EmilienM | mtreinish: how grenade validates upgrades don't break resources? | 20:42 |
| clarkb | mtreinish: I think you want to close them when you are done with them as a result too | 20:43 |
| mtreinish | EmilienM: I thought we checked with you multiple times before we removed javelin from the tempest repo | 20:43 |
| mtreinish | clarkb: I think it does that already | 20:43 |
| mtreinish | just in an outside function | 20:43 |
| mtreinish | clarkb: but we could still be leaking elsewhere | 20:43 |
| EmilienM | mtreinish: wait, i'm not complaining :) | 20:43 |
| mtreinish | EmilienM: heh, ok :) | 20:43 |
| EmilienM | mtreinish: I'm just wondering how it works now | 20:43 |
| EmilienM | mtreinish: in upgrade testing | 20:43 |
| EmilienM | because I want to reuse it for tripleo CI | 20:44 |
| mtreinish | EmilienM: so grenade has resource scripts per project: https://github.com/openstack-dev/grenade/blob/master/projects/60_nova/resources.sh | 20:44 |
| clarkb | mtreinish: looks like it just failed in the gate fwiw | 20:44 |
| EmilienM | that ! | 20:44 |
| EmilienM | mtreinish: I'll try to re-use this code | 20:44 |
| EmilienM | mtreinish: ty sir | 20:44 |
| mtreinish | clarkb: it did, but it shows the real underlying failure, which is related to metadata getting in the guest | 20:45 |
| mtreinish | clarkb: instead of the banner error | 20:45 |
| clarkb | oh good | 20:45 |
| clarkb | metadata errors are my favorite | 20:45 |
| openstackrecheck | Console logs not available after 20:16s for gate-group-based-policy-python27-ubuntu-xenial 426437,35,c2f3dc9 | 20:45 |
| mtreinish | clarkb: because it failed since the pubkey was rejected on all the login attempts | 20:45 |
| mtreinish | clarkb: which is what I was pointing kevinbenton and frickler at :) | 20:45 |
| mtreinish | EmilienM: sure, np | 20:45 |
| clarkb | yay real bugs | 20:45 |
| *** SlickN1k has joined #openstack-qa | 20:46 | |
| mtreinish | clarkb: I was going to send a follow up to our state of the gate email about this after the close() patch merged | 20:47 |
| *** porrua has quit IRC | 20:47 | |
| *** stream10 has quit IRC | 20:47 | |
| clarkb | mtreinish: cool | 20:48 |
| *** awaugama has quit IRC | 20:49 | |
| *** beisner- has joined #openstack-qa | 20:50 | |
| *** mrodden_ has joined #openstack-qa | 20:51 | |
| *** mmedvede_ has joined #openstack-qa | 20:52 | |
| *** mmedvede has quit IRC | 20:53 | |
| *** sirushti has quit IRC | 20:53 | |
| *** mrodden has quit IRC | 20:53 | |
| *** SlickNik has quit IRC | 20:53 | |
| *** beisner has quit IRC | 20:53 | |
| *** beisner- is now known as beisner | 20:53 | |
| *** mrodden_ is now known as mrodden | 20:53 | |
| *** mmedvede_ is now known as mmedvede | 20:53 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck master: Use modern PBR and hacking packages https://review.openstack.org/439844 | 20:53 |
| mtreinish | sdague: that was quick :) | 20:54 |
| *** Apoorva has quit IRC | 20:55 | |
| *** sirushti has joined #openstack-qa | 20:55 | |
| sdague | meh, why wait :) | 20:56 |
| mtreinish | there was no reason, you just +A'd it before I even opened up the review in my browser | 20:57 |
| mtreinish | but, I blame the poor wifi at the starbucks by marist for that more than anything | 20:58 |
| sdague | heh | 20:59 |
| *** Apoorva has joined #openstack-qa | 20:59 | |
| *** harlowja has joined #openstack-qa | 21:01 | |
| *** harlowja has quit IRC | 21:01 | |
| *** harlowja has joined #openstack-qa | 21:01 | |
| *** pkoniszewski has quit IRC | 21:02 | |
| openstackgerrit | Merged openstack-infra/elastic-recheck master: Use modern PBR and hacking packages https://review.openstack.org/439844 | 21:02 |
| openstackrecheck | Console logs not available after 20:17s for gate-telemetry-dsvm-integration-ceilometer-ubuntu-xenial 438338,1,af108b0 | 21:05 |
| *** Nakato has quit IRC | 21:06 | |
| *** Nakato has joined #openstack-qa | 21:07 | |
| *** sirushti has quit IRC | 21:07 | |
| *** sirushti has joined #openstack-qa | 21:08 | |
| *** catintheroof has joined #openstack-qa | 21:08 | |
| *** pkoniszewski has joined #openstack-qa | 21:09 | |
| *** krtaylor has joined #openstack-qa | 21:09 | |
| *** yamamoto has joined #openstack-qa | 21:10 | |
| *** DotanSoko has quit IRC | 21:10 | |
| *** catintheroof has quit IRC | 21:10 | |
| *** edannon has quit IRC | 21:10 | |
| *** catintheroof has joined #openstack-qa | 21:10 | |
| *** pkoniszewski has quit IRC | 21:12 | |
| *** yamamoto has quit IRC | 21:12 | |
| *** yamamoto has joined #openstack-qa | 21:12 | |
| *** openstackrecheck has quit IRC | 21:13 | |
| *** openstackrecheck has joined #openstack-qa | 21:13 | |
| *** dimtruck is now known as zz_dimtruck | 21:14 | |
| sdague | mtreinish / andreaf - when is the last time you got a fully green tempest run on the first go - https://review.openstack.org/#/c/439698/ :) | 21:14 |
| *** salv-orlando has joined #openstack-qa | 21:14 | |
| mtreinish | sdague: this morning | 21:15 |
| sdague | mtreinish: I think it's been six months for me | 21:16 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck master: Add query for bug 1668958 https://review.openstack.org/439562 | 21:16 |
| openstack | bug 1668958 in OpenStack Compute (nova) "metadata service occasionally not returning keys" [Undecided,New] https://launchpad.net/bugs/1668958 | 21:16 |
| *** raildo has quit IRC | 21:17 | |
| mtreinish | kevinbenton: ^^^ I updated the query | 21:17 |
| sdague | mtreinish: you mean the one that just failed the gate? - https://review.openstack.org/#/c/439638/ ;) | 21:18 |
| mtreinish | it was all green on the first go | 21:18 |
| sdague | I'm confused, did it fail 1 or 2 tests? I only see the one in the dump | 21:19 |
| sdague | oh, I see the second one now | 21:19 |
| *** pkoniszewski has joined #openstack-qa | 21:19 | |
| sdague | the differentiation isn't great | 21:19 |
| *** yamamoto has quit IRC | 21:19 | |
| *** pkoniszewski has quit IRC | 21:22 | |
| *** vaidy has joined #openstack-qa | 21:28 | |
| *** stream10 has joined #openstack-qa | 21:30 | |
| *** pcaruana has joined #openstack-qa | 21:30 | |
| *** pcaruana has quit IRC | 21:33 | |
| *** Serlex has quit IRC | 21:33 | |
| *** gouthamr has quit IRC | 21:33 | |
| *** gcheresh_ has quit IRC | 21:33 | |
| openstackrecheck | Console logs not available after 20:15s for gate-tempest-dsvm-neutron-linuxbridge-ubuntu-xenial 439076,3,57d236d | 21:38 |
| *** gyee has joined #openstack-qa | 21:40 | |
| *** aysyd has quit IRC | 21:42 | |
| *** jdandrea has quit IRC | 21:43 | |
| *** pvaneck has joined #openstack-qa | 21:43 | |
| openstackgerrit | Merged openstack/tempest master: Adds missing server tags APIs to servers client. https://review.openstack.org/435630 | 21:44 |
| openstackgerrit | fumihiko kakuma proposed openstack-dev/devstack master: DO NOT MERGE: This is to check ryu master https://review.openstack.org/439383 | 21:50 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Move CONF values into __init__ in RemoteClient https://review.openstack.org/439863 | 21:50 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Add dhcp_client argument to renew_lease() https://review.openstack.org/439864 | 21:50 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Add network_mask_bits argument to assign_static_ip() https://review.openstack.org/439865 | 21:50 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Outout server id only on ssh debugging https://review.openstack.org/439866 | 21:50 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: WIP: Add RemoteClient under tempest.lib https://review.openstack.org/439867 | 21:50 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Deprecate resources_prefix and change rand_name() https://review.openstack.org/439203 | 21:57 |
| *** boris-42_ has quit IRC | 22:01 | |
| *** jdandrea has joined #openstack-qa | 22:01 | |
| *** Apoorva_ has joined #openstack-qa | 22:02 | |
| *** ihrachys has quit IRC | 22:03 | |
| *** Apoorva_ has quit IRC | 22:03 | |
| *** Apoorva has quit IRC | 22:03 | |
| *** Apoorva has joined #openstack-qa | 22:03 | |
| *** yamamoto has joined #openstack-qa | 22:05 | |
| andreaf | sdague: thank for putting https://review.openstack.org/#/c/439698/2 up | 22:06 |
| *** xinli has quit IRC | 22:07 | |
| *** dsariel has joined #openstack-qa | 22:09 | |
| *** xinli has joined #openstack-qa | 22:09 | |
| openstackrecheck | Console logs not available after 20:18s for gate-rally-dsvm-neutron-neutron-ubuntu-xenial 439814,1,ec9fe88 | 22:10 |
| *** stream10 has quit IRC | 22:14 | |
| andreaf | sdague: comparing dtsat graphs it looks perhaps slightly better, but not dramatically - on CPU, free memory, load avg | 22:14 |
| andreaf | sdague: I think the API tests do a rather good job in keeping the system busy, even more than the scenario ones | 22:16 |
| andreaf | sdague: scenario tests tend to be scheduled towards the end of the test run, and with your change the last part of the graph looks a bit flatter | 22:17 |
| openstackgerrit | Rick Bartra proposed openstack/patrole master: Compute API Keypairs https://review.openstack.org/439877 | 22:20 |
| *** gyee has quit IRC | 22:20 | |
| *** dave-mccowan has quit IRC | 22:21 | |
| *** zz_dimtruck is now known as dimtruck | 22:22 | |
| *** armax has quit IRC | 22:26 | |
| openstackgerrit | Andrea Frittoli proposed openstack/tempest master: DNM Set concurrency to 3 https://review.openstack.org/439879 | 22:26 |
| *** dsariel has quit IRC | 22:27 | |
| *** dustins_ has quit IRC | 22:31 | |
| *** felipemonteiro__ has quit IRC | 22:32 | |
| *** felipemonteiro_ has quit IRC | 22:32 | |
| *** yangyapeng has quit IRC | 22:33 | |
| *** gyee has joined #openstack-qa | 22:40 | |
| *** jdandrea has quit IRC | 22:42 | |
| *** sshnaidm is now known as sshnaidm|off | 22:44 | |
| *** breitz_ has joined #openstack-qa | 22:45 | |
| *** breitz has quit IRC | 22:46 | |
| openstackrecheck | Console logs not available after 20:21s for gate-tempest-dsvm-neutron-full-ubuntu-xenial 435810,6,fe3fe06 | 22:49 |
| *** salv-orlando has quit IRC | 22:50 | |
| *** salv-orlando has joined #openstack-qa | 22:51 | |
| openstackgerrit | Rodrigo Duarte proposed openstack-dev/devstack master: Add allow_global_implied_dsr_disabled feature flag https://review.openstack.org/428292 | 22:51 |
| *** breitz has joined #openstack-qa | 22:52 | |
| *** breitz_ has quit IRC | 22:54 | |
| openstackgerrit | Rodrigo Duarte proposed openstack-dev/devstack master: Add allow_global_implied_dsr_disabled feature flag https://review.openstack.org/428292 | 22:54 |
| openstackgerrit | Rodrigo Duarte proposed openstack/tempest master: Improve test_implied_domain_roles https://review.openstack.org/427345 | 22:55 |
| *** tonytan4ever has quit IRC | 22:57 | |
| *** gouthamr has joined #openstack-qa | 22:58 | |
| openstackgerrit | John L. Villalovos proposed openstack-dev/grenade master: WIP/DNM: Debugging https://review.openstack.org/439893 | 22:58 |
| openstackgerrit | Alex Schultz proposed openstack-infra/elastic-recheck master: Add recheck rule for RDO package download failures https://review.openstack.org/439895 | 23:00 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: WIP: Add RemoteClient under tempest.lib https://review.openstack.org/439867 | 23:10 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Remove get_ram_size_in_mb, get_number_of_vcpus, etc https://review.openstack.org/439899 | 23:10 |
| *** fnaval has quit IRC | 23:11 | |
| openstackrecheck | Console logs not available after 20:09s for gate-dragonflow-requirements 424146,9,3cfd9c4 | 23:12 |
| *** csomerville has quit IRC | 23:16 | |
| *** cody-somerville has joined #openstack-qa | 23:17 | |
| *** cody-somerville has quit IRC | 23:17 | |
| *** cody-somerville has joined #openstack-qa | 23:17 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Outout server id only on ssh debugging https://review.openstack.org/439866 | 23:18 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: WIP: Add RemoteClient under tempest.lib https://review.openstack.org/439867 | 23:18 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/tempest master: Remove get_ram_size_in_mb, get_number_of_vcpus, etc https://review.openstack.org/439899 | 23:18 |
| *** gyee has quit IRC | 23:18 | |
| *** gyee has joined #openstack-qa | 23:19 | |
| *** rfolco has quit IRC | 23:22 | |
| *** mriedem has quit IRC | 23:24 | |
| openstackgerrit | Vasyl Saienko proposed openstack-dev/grenade master: Apply OS_TEST_TIMEOUT when running tempest https://review.openstack.org/438955 | 23:29 |
| *** catintheroof has quit IRC | 23:29 | |
| openstackrecheck | Console logs not available after 20:10s for gate-dragonflow-dsvm-fullstack-ubuntu-xenial 439848,1,3ef05cf | 23:32 |
| *** Apoorva_ has joined #openstack-qa | 23:32 | |
| *** salv-orlando has quit IRC | 23:35 | |
| *** Apoorva has quit IRC | 23:36 | |
| *** rbak has quit IRC | 23:36 | |
| openstackgerrit | fumihiko kakuma proposed openstack-dev/devstack master: DO NOT MERGE: This is to check ryu master https://review.openstack.org/439383 | 23:42 |
| *** dimtruck is now known as zz_dimtruck | 23:45 | |
| openstackgerrit | Anthony Bellino proposed openstack/patrole master: Orchestration API config tests https://review.openstack.org/439916 | 23:46 |
| *** dave-mccowan has joined #openstack-qa | 23:51 | |
| *** dave-mccowan has quit IRC | 23:51 | |
| openstackrecheck | Console logs not available after 20:14s for gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial 439156,2,0b8a23e | 23:54 |
| *** vhosakot has quit IRC | 23:54 | |
| *** zz_dimtruck is now known as dimtruck | 23:56 | |
| *** Guest27057 is now known as zigo | 23:57 | |
| *** xinli has quit IRC | 23:57 | |
| *** sdague has quit IRC | 23:58 | |
| gmann | sdague: andreaf yea i completely agree on cutting out the scenario tests on gate. | 23:58 |
| gmann | sdague: andreaf even there are lot of tests in scenario which does not make sense there. i will try to cleanup those. | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!