Friday, 2015-04-17

*** mahito has joined #openstack-qa00:01
openstackgerritJoe Gordon proposed openstack/tempest: Skip test_shelve_instance in multihost case  https://review.openstack.org/17465200:02
jogomtreinish: something like that ^00:02
mtreinishjogo: yeah that should work00:03
*** mahito_ has joined #openstack-qa00:04
*** dims__ has joined #openstack-qa00:04
mtreinishjogo: after looking at that test, your the other option would be to tell devstack to say shelve doesn't work in multinode00:04
mtreinishbut I think we should save that if we figure out it really doesn't work multinode00:05
*** mahito has quit IRC00:05
*** dims_ has joined #openstack-qa00:05
jogoyeah this should be a bug00:06
jogonot a issue we just have to live with00:07
jogomtreinish: thanks00:07
* jogo wanders off00:08
*** dims__ has quit IRC00:09
*** bknudson has joined #openstack-qa00:12
*** davideagnello has quit IRC00:12
*** tosky has quit IRC00:13
*** sputnik13 has quit IRC00:13
*** luqas has joined #openstack-qa00:15
*** davideagnello has joined #openstack-qa00:16
*** luqas has quit IRC00:20
*** dmellado has quit IRC00:21
openstackgerritMerged openstack-dev/devstack: doc changes to devstack overview.rst  https://review.openstack.org/17449600:22
*** yamamoto has quit IRC00:25
*** dmorita has joined #openstack-qa00:27
*** yamamoto has joined #openstack-qa00:27
*** yamamoto has quit IRC00:28
*** dmellado has joined #openstack-qa00:29
*** marun has quit IRC00:31
*** davideagnello has quit IRC00:32
*** zz_dimtruck is now known as dimtruck00:34
*** mlavalle has quit IRC00:37
openstackgerritIan Wienand proposed openstack-dev/devstack: Actually run function tests  https://review.openstack.org/17466300:41
ianwdougwig: ^ so we don't even run those tests in the gate!  doh00:42
openstackgerritIan Wienand proposed openstack-dev/devstack: Stop deleting token prefixes when processing ENABLED_SERVICES negations  https://review.openstack.org/17140200:43
dougwigianw: ooh, good find.00:43
ianwdougwig: i just did a depend-on for your change, that should actually test it :)00:44
clarkbjogo mtreinish shelve does work across nodes00:44
clarkbI worked hard to make that work00:44
clarkbif it doesnt work that is a bug likely in nova00:44
clarkbback to driving now but ^ is a regression if thats the case. now is time to fix it not ignore it00:46
mtreinishclarkb: ok well I'll drop my +2 on jogo's skip commit then00:46
mtreinishhe made it sound like a fix wasn't coming any time soon and a skip was necessary00:47
*** patrickeast has quit IRC00:50
openstackgerritMatthew Treinish proposed openstack/tempest: Cleanup fixed_name logic in test_list_server_filters  https://review.openstack.org/17467100:50
mtreinishsdague: ^^^00:50
openstackgerritDoug Wiegley proposed openstack-dev/devstack: Stop deleting token prefixes when processing ENABLED_SERVICES negations  https://review.openstack.org/17140200:51
*** achanda has quit IRC00:51
*** achanda has joined #openstack-qa00:59
*** mwagner_lap has joined #openstack-qa01:00
*** weshay has joined #openstack-qa01:05
openstackgerritMatthew Treinish proposed openstack/tempest: Add a paragraph on the use of tempest_roles option  https://review.openstack.org/17467501:05
*** yamamoto has joined #openstack-qa01:10
*** krtaylor has joined #openstack-qa01:13
clarkbmtreinish why wouldnt we fix it during bug fix time?01:15
clarkbalso note I have been afk all day so may have missed something01:15
clarkbpreviously the issue was shelve wouldnt cleanup networking on old host then bring it up on new host so packet paths were messed up01:17
clarkband it was a real bug01:18
*** jecarey has joined #openstack-qa01:25
*** Sukhdev has quit IRC01:34
*** rfolco has quit IRC01:46
*** harlowja is now known as harlowja_away01:47
*** dims_ has quit IRC01:54
openstackgerritIan Wienand proposed openstack-dev/devstack: Create config file in merge_config_file  https://review.openstack.org/17467801:57
ianwdougwig: ^ another thing i noticed01:57
ianwdougwig: also, https://jenkins07.openstack.org/job/gate-devstack-unit-tests/393/console <- your change does seem to fail, but we don't pick it up01:58
openstackrecheckConsole logs not available after 13:22s for gate-sahara-pep8 173654,11,339252f02:08
*** pal has joined #openstack-qa02:13
*** achanda has quit IRC02:20
*** dkranz has joined #openstack-qa02:21
*** achanda has joined #openstack-qa02:22
*** dmellado has quit IRC02:22
*** pal_ has joined #openstack-qa02:33
*** pal has quit IRC02:34
*** dimtruck is now known as zz_dimtruck02:34
*** marrusl has quit IRC02:34
openstackgerritKen'ichi Ohmichi proposed openstack/tempest: Check domain's enabled attribute strictly  https://review.openstack.org/17468002:35
*** dkranz has quit IRC02:37
*** weshay has quit IRC02:37
*** dmellado has joined #openstack-qa02:42
*** liusheng has joined #openstack-qa02:46
*** jecarey has quit IRC02:47
*** achanda has quit IRC02:49
*** woodster_ has quit IRC02:50
*** sputnik13 has joined #openstack-qa02:53
*** yamahata has quit IRC02:56
*** Longgeek has joined #openstack-qa02:59
*** Longgeek has quit IRC03:04
*** Longgeek has joined #openstack-qa03:06
*** liusheng has quit IRC03:07
*** liusheng has joined #openstack-qa03:07
*** achanda has joined #openstack-qa03:20
*** pal_ has quit IRC03:22
openstackgerritIan Wienand proposed openstack-dev/devstack: Create config file in merge_config_file  https://review.openstack.org/17467803:24
openstackgerritIan Wienand proposed openstack-dev/devstack: Add "passed" and "failed" functions  https://review.openstack.org/17468203:24
openstackgerritIan Wienand proposed openstack-dev/devstack: Use unittest help in test_ip.sh  https://review.openstack.org/17468303:24
openstackgerritIan Wienand proposed openstack-dev/devstack: Fix return of test_ini_config & test_meta_config  https://review.openstack.org/17468403:24
openstackgerritIan Wienand proposed openstack-dev/devstack: Move function.sh to test_functions.sh  https://review.openstack.org/17468503:24
openstackgerritIan Wienand proposed openstack-dev/devstack: Fix die_if_not_set tests  https://review.openstack.org/17468603:24
openstackgerritIan Wienand proposed openstack-dev/devstack: Convert test_functions.sh to use unittest helpers  https://review.openstack.org/17468703:24
openstackgerritIan Wienand proposed openstack-dev/devstack: Remove old comment in run_tests.sh  https://review.openstack.org/17468803:24
ianwdougwig: ^ you see what you created :)03:24
*** ebagdasa has quit IRC03:29
*** pal has joined #openstack-qa03:31
*** davideagnello has joined #openstack-qa03:33
*** davideagnello has quit IRC03:38
*** Sukhdev has joined #openstack-qa03:44
dougwigianw: good, this is more valuable than the original change.  ðŸ˜ƒ03:44
*** achanda has quit IRC03:52
*** luqas has joined #openstack-qa03:52
*** sputnik13 has quit IRC03:53
*** luqas has quit IRC03:57
dougwigianw: anything I can help with in the morning?04:00
ianwdougwig: i don't think so; that stack mean that ./run_tests.sh runs the tests and catches failures.  but i do believe your proposed original change fails, despite not being picked up ATM04:01
dougwigianw: I saw that, and marked my patch -a to make sure it's not considered.04:02
ianwdougwig: also maybe double-check https://review.openstack.org/#/c/174678/ -- i get so lost in that ridiculously long awk script ...04:02
dougwigianw: awk, perhaps the only thing that can make perl look good.04:04
*** yfried|afk is now known as yfried__04:06
*** achanda has joined #openstack-qa04:07
*** achanda has quit IRC04:11
*** achanda has joined #openstack-qa04:16
*** yfried__ is now known as yfried|afk04:17
*** achanda has quit IRC04:18
*** yfried|afk is now known as yfried__04:19
*** achanda has joined #openstack-qa04:21
openstackgerritMerged openstack/tempest: Test creation of server attached to created port  https://review.openstack.org/16288604:22
*** achanda has quit IRC04:29
*** kashyap` is now known as kashyap04:36
*** achanda has joined #openstack-qa04:42
*** mmedvede has quit IRC04:46
*** ajmiller_ has joined #openstack-qa04:51
*** neeti has joined #openstack-qa04:52
*** ajmiller has quit IRC04:54
*** mmedvede has joined #openstack-qa04:57
*** oomichi has joined #openstack-qa05:08
*** sputnik13 has joined #openstack-qa05:09
openstackgerritAbhishek G M proposed openstack/tempest: Test Creation of server in a shared_network  https://review.openstack.org/16891005:11
*** sputnik13 has quit IRC05:11
openstackgerritAbhishek G M proposed openstack/tempest: Test Create of VM with multiple Nics  https://review.openstack.org/16804005:20
*** openstackgerrit has quit IRC05:21
*** openstackgerrit has joined #openstack-qa05:21
*** sks has joined #openstack-qa05:27
*** Madan has joined #openstack-qa05:27
*** yfried__ is now known as yfried|afk05:34
*** rushiagr_away is now known as rushiagr05:38
*** nikil2 has joined #openstack-qa05:42
*** hitalia has joined #openstack-qa05:43
*** Sukhdev has quit IRC05:43
*** Sukhdev has joined #openstack-qa05:50
*** oomichi has quit IRC05:59
*** pal has quit IRC06:10
*** ibiris_away is now known as ibiris06:17
*** armax has quit IRC06:22
*** bnemec has quit IRC06:32
*** k4n0 has joined #openstack-qa06:34
*** Longgeek_ has joined #openstack-qa06:37
*** yamahata has joined #openstack-qa06:38
*** Longgeek has quit IRC06:39
*** jyuso1 has quit IRC06:40
*** mhorban has quit IRC06:45
*** Sukhdev has quit IRC06:45
*** achanda has quit IRC06:53
*** achanda has joined #openstack-qa06:56
*** yfried|afk is now known as yfried__06:56
*** gilliard is now known as gillllliard07:03
*** gillllliard is now known as gillllllliard07:03
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Add support for provider network in tempest  https://review.openstack.org/15688707:06
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Add support for provider network in tempest  https://review.openstack.org/17472907:07
*** e0ne has joined #openstack-qa07:08
openstackgerritMh Raies proposed openstack/tempest-lib: Adding support for trove CLI.  https://review.openstack.org/17473007:08
*** yfried__ is now known as yfried|afk07:12
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Fix issue with ml2 plugin on using provider network  https://review.openstack.org/17473107:13
*** markus_z has joined #openstack-qa07:14
*** ibiris is now known as ibiris_away07:14
*** ibiris_away is now known as ibiris07:15
*** amotoki has quit IRC07:18
*** svasheka has quit IRC07:21
*** nikil2 has quit IRC07:21
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Add support for ilo drivers in lib/ironic  https://review.openstack.org/17473307:21
*** luqas has joined #openstack-qa07:22
*** e0ne has quit IRC07:23
*** tochi has quit IRC07:23
*** e0ne has joined #openstack-qa07:28
*** tochi has joined #openstack-qa07:28
*** ibiris is now known as ibiris_away07:28
*** achanda has quit IRC07:29
*** andreykurilin__ has joined #openstack-qa07:32
*** achanda_ has joined #openstack-qa07:32
*** dims__ has joined #openstack-qa07:33
*** yamahata has quit IRC07:34
*** jlibosva has joined #openstack-qa07:35
*** nikil22 has joined #openstack-qa07:36
*** vponomaryov has quit IRC07:37
*** dims__ has quit IRC07:38
*** jlanoux has joined #openstack-qa07:40
*** achanda_ has quit IRC07:40
openstackrecheckConsole logs not available after 13:25s for gate-cinder-docs 174482,2,56b980407:40
*** gfidente has joined #openstack-qa07:43
*** yfried|afk is now known as yfried__07:44
*** mahito has joined #openstack-qa07:46
*** alex_xu has quit IRC07:46
*** heyongli has quit IRC07:47
*** heyongli has joined #openstack-qa07:48
*** alex_xu has joined #openstack-qa07:48
*** mahito_ has quit IRC07:49
*** andreykurilin___ has joined #openstack-qa07:53
*** andreykurilin__ has quit IRC07:53
*** e0ne has quit IRC07:57
*** ibiris_away is now known as ibiris08:01
*** e0ne has joined #openstack-qa08:01
*** SridharG has joined #openstack-qa08:04
*** rushiagr is now known as rushiagr_away08:04
*** achanda has joined #openstack-qa08:06
*** sks has quit IRC08:08
*** e0ne has quit IRC08:09
openstackgerritMh Raies proposed openstack/tempest-lib: Adding support for trove CLI.  https://review.openstack.org/17473008:10
*** newbie has joined #openstack-qa08:12
*** newbie is now known as Guest4383608:13
*** e0ne has joined #openstack-qa08:14
*** e0ne has quit IRC08:18
*** turul has joined #openstack-qa08:18
*** achanda has quit IRC08:20
*** hichihara has quit IRC08:21
*** ylobankov_ has quit IRC08:21
*** achanda_ has joined #openstack-qa08:22
*** zz_johnthetubagu is now known as johnthetubaguy08:22
*** yfried__ is now known as yfried|afk08:22
*** e0ne has joined #openstack-qa08:22
Guest43836Could anyone take a look, please: https://review.openstack.org/17387508:26
*** ylobankov has joined #openstack-qa08:26
*** safchain has joined #openstack-qa08:29
*** turul is now known as afazekas08:30
*** achanda_ has quit IRC08:35
*** mahito has quit IRC08:37
*** yfried|afk is now known as yfried__08:37
*** jordanP has joined #openstack-qa08:42
*** e0ne has quit IRC08:42
*** andreykurilin___ has quit IRC08:43
*** e0ne has joined #openstack-qa08:46
*** pcaruana has joined #openstack-qa08:50
*** mpaolino has joined #openstack-qa08:50
*** e0ne has quit IRC08:51
*** e0ne has joined #openstack-qa08:53
*** tkammer has joined #openstack-qa08:58
*** e0ne has quit IRC09:00
*** pal has joined #openstack-qa09:00
*** e0ne has joined #openstack-qa09:04
*** yamamoto has quit IRC09:06
*** e0ne has quit IRC09:09
*** ibiris is now known as ibiris_away09:09
*** rushiagr_away is now known as rushiagr09:10
*** yfried__ is now known as yfried|afk09:12
*** yfried|afk is now known as yfried__09:15
*** ihrachyshka has joined #openstack-qa09:16
*** gszasz has joined #openstack-qa09:18
*** ibiris_away is now known as ibiris09:19
*** anshul has joined #openstack-qa09:22
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Add support for ilo drivers in lib/ironic  https://review.openstack.org/17473309:31
*** ibiris is now known as ibiris_away09:33
*** yfried__ is now known as yfried|afk09:34
*** ibiris_away is now known as ibiris09:35
openstackrecheckConsole logs not available after 13:23s for gate-nova-pep8 174656,6,0b8267709:43
*** aix has joined #openstack-qa09:45
*** jamielennox is now known as jamielennox|away09:49
*** e0ne has joined #openstack-qa09:50
*** tosky has joined #openstack-qa09:54
*** cdent has joined #openstack-qa10:01
*** enikanorov_ has quit IRC10:09
*** sputnik13 has joined #openstack-qa10:12
*** cdent has quit IRC10:12
*** cdent has joined #openstack-qa10:13
*** enikanorov_ has joined #openstack-qa10:15
*** sputnik13 has quit IRC10:15
*** neeti has quit IRC10:17
*** yfried|afk is now known as yfried__10:22
*** pc_m has joined #openstack-qa10:28
*** Madan has quit IRC10:30
*** safchain has quit IRC10:31
*** safchain has joined #openstack-qa10:34
*** e0ne is now known as e0ne_10:38
*** yfried__ is now known as yfried|afk10:38
*** dratushnyy has quit IRC10:38
*** ajmiller__ has joined #openstack-qa10:39
openstackgerritfumihiko kakuma proposed openstack-dev/devstack: Fix remove_disabled_extensions to remove an extension at the last position  https://review.openstack.org/17278910:40
*** e0ne_ is now known as e0ne10:42
*** ajmiller_ has quit IRC10:43
*** pal has quit IRC10:45
*** k4n0 has quit IRC10:47
*** jlibosva has quit IRC10:53
*** jlibosva has joined #openstack-qa10:54
*** jlibosva has quit IRC10:54
*** jlibosva has joined #openstack-qa10:55
*** jlibosva has quit IRC10:56
*** jlibosva has joined #openstack-qa10:56
*** yfried|afk is now known as yfried__10:57
*** jlibosva has quit IRC10:57
*** jlibosva has joined #openstack-qa10:57
*** dims__ has joined #openstack-qa11:02
*** sks has joined #openstack-qa11:05
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Add support for ilo drivers in lib/ironic  https://review.openstack.org/17473311:05
*** aix has quit IRC11:05
*** vponomaryov has joined #openstack-qa11:06
*** yfried__ is now known as yfried|afk11:07
*** yfried|afk is now known as yfried__11:09
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Configure active_timeout and ssh_user in lib/tempest  https://review.openstack.org/17481911:17
openstackgerritRamakrishnan G proposed openstack/tempest: Make OS_TEST_TIMEOUT configurable in tox.ini  https://review.openstack.org/17482211:23
raies_afazekas: hi11:26
*** mwagner_lap has quit IRC11:26
raies_afazekas: I am facing an error, Jenkins gets fail because "Cells" is not enabled. Is by default n-cells service is disabled in devstack-gate ? how to enable it ? I can see default setting to disable here https://github.com/openstack-infra/devstack-gate/blob/master/devstack-vm-gate-wrap.sh#L20411:28
openstackgerritRamakrishnan G proposed openstack/tempest: Honour ssh_connect_method in test_baremetal_server_ops  https://review.openstack.org/17482811:34
openstackgerritSean Dague proposed openstack-dev/devstack: refactor ping_check  https://review.openstack.org/17437011:39
*** dmorita has quit IRC11:44
openstackrecheckConsole logs not available after 13:25s for gate-nova-pep8 118604,17,ef8ffd311:45
afazekasraies_,  We had a non voting cell job only on nova and it is failing: http://logs.openstack.org/94/173794/2/check/check-tempest-dsvm-cells/ec1ae52/11:47
*** tochi has quit IRC11:48
afazekass/had/have/11:49
*** anshul has quit IRC11:49
*** anshul has joined #openstack-qa11:50
raies_afazekas: I faced http://logs.openstack.org/48/171148/2/check/check-novaclient-dsvm-functional/273fef7/testr_results.html.gz11:50
afazekasraies_: IMHO if we would like to have cell job on tempest the layout in the project-config repo needs to be changed11:50
*** anshul has quit IRC11:50
*** anshul has joined #openstack-qa11:51
raies_afazekas: ok11:51
openstackgerritRamakrishnan G proposed openstack-dev/devstack: Network not reachable after creating provider net  https://review.openstack.org/17483211:51
*** cbits has joined #openstack-qa11:52
*** cbits has left #openstack-qa11:55
*** cdent has quit IRC11:55
*** e0ne is now known as e0ne_11:58
openstackrecheckConsole logs not available after 13:22s for gate-cinder-pep8 163706,7,9195fea11:59
*** cdent has joined #openstack-qa12:01
*** aix has joined #openstack-qa12:03
*** rfolco has joined #openstack-qa12:06
*** aysyd has joined #openstack-qa12:06
*** e0ne_ has quit IRC12:08
*** e0ne has joined #openstack-qa12:10
*** weshay has joined #openstack-qa12:13
*** bknudson has quit IRC12:15
*** mwagner_lap has joined #openstack-qa12:19
*** woodster_ has joined #openstack-qa12:19
*** abehl has joined #openstack-qa12:20
*** anshul has quit IRC12:21
openstackgerritSean Dague proposed openstack-dev/devstack: add back n-cert  https://review.openstack.org/17484712:23
*** sks has quit IRC12:25
*** yfried__ is now known as yfried|afk12:29
*** yfried|afk is now known as yfried__12:29
*** cdent has quit IRC12:34
*** cdent has joined #openstack-qa12:35
*** bknudson has joined #openstack-qa12:36
*** jaypipes has joined #openstack-qa12:38
*** ChanServ sets mode: +o jaypipes12:38
*** cdent has quit IRC12:41
*** cdent has joined #openstack-qa12:41
*** nikil22 has quit IRC12:44
*** dpaterson has joined #openstack-qa12:49
*** anshul has joined #openstack-qa12:50
*** dkranz has joined #openstack-qa12:51
*** abehl has quit IRC12:52
*** cbader has joined #openstack-qa12:53
*** SridharG has left #openstack-qa12:56
*** zz_dimtruck is now known as dimtruck12:59
*** cbader has quit IRC13:00
*** marrusl has joined #openstack-qa13:01
*** dmellado has quit IRC13:02
openstackgerritSean Dague proposed openstack-dev/devstack: add back n-cert  https://review.openstack.org/17484713:02
*** neeti has joined #openstack-qa13:03
*** neeti has quit IRC13:03
*** gszasz has quit IRC13:06
*** artom has joined #openstack-qa13:14
*** marcusvrn has quit IRC13:19
*** cdent has quit IRC13:23
openstackgerritAnthony PERARD proposed openstack/tempest: Introduce compute.libvirt_type config option.  https://review.openstack.org/17486613:23
openstackgerritAnthony PERARD proposed openstack/tempest: Skip test_volume_boot_pattern for Nova with libvirt-xen  https://review.openstack.org/17486713:23
*** dustins has joined #openstack-qa13:23
*** rushiagr is now known as rushiagr_away13:23
*** mriedem has joined #openstack-qa13:26
*** markus_z has quit IRC13:29
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add a cli subcommand for deleting old rows  https://review.openstack.org/17220013:29
*** jecarey has joined #openstack-qa13:32
*** dansmith is now known as superdan13:35
*** yfried__ is now known as yfried|afk13:41
*** yfried|afk is now known as yfried__13:43
*** anshul has quit IRC13:45
*** anshul has joined #openstack-qa13:46
*** anshul has quit IRC13:46
*** anshul has joined #openstack-qa13:47
*** mhorban has joined #openstack-qa13:49
openstackgerritDan Smith proposed openstack-dev/grenade: DNM: For testing nova  https://review.openstack.org/17458113:50
openstackgerritDan Smith proposed openstack-dev/grenade: Make nova migrate flavor data, per Kilo->Lemming requirements  https://review.openstack.org/17456713:50
*** mpaolino has quit IRC13:53
openstackgerritThomas Bechtold proposed openstack/tempest-lib: Fix ValueError in subunit_trace  https://review.openstack.org/17488213:57
*** dimtruck is now known as zz_dimtruck13:58
*** jaypipes is now known as leakypipes13:59
*** rbak has joined #openstack-qa14:00
*** bnemec has joined #openstack-qa14:03
*** mriedem has quit IRC14:03
*** yfried__ is now known as yfried|afk14:03
*** mriedem has joined #openstack-qa14:06
*** BadCub_Hiding is now known as BadCub14:06
openstackgerritSergey Slipushenko proposed openstack/tempest: Tool for generation user accounts from spec  https://review.openstack.org/17260514:07
*** zz_dimtruck is now known as dimtruck14:07
*** rushiagr_away is now known as rushiagr14:08
*** cdent has joined #openstack-qa14:09
*** yfried|afk is now known as yfried__14:09
sslypush1nkomtreinish: I separated account_generator code from javelin, as you asked me to do. Can you review this patch ^^14:10
openstackrecheckopenstack-dev/devstack change: https://review.openstack.org/173493 failed because of: gate-tempest-dsvm-full: https://bugs.launchpad.net/bugs/142915114:10
openstackLaunchpad bug 1429151 in Cinder "SimpleReadOnlyCinderClientTest.test_cinder_encryption_type_list fails randomly" [High,Confirmed]14:10
mtreinishsslypush1nko: yes I will take a look14:10
openstackgerritAnthony PERARD proposed openstack-dev/devstack: Include libvirt_type in tempest config.  https://review.openstack.org/17488814:10
sslypush1nkothx14:10
*** anshul has quit IRC14:11
sslypush1nkoAlso I left feature for cleanup, becuse it is in a spec14:11
*** e0ne is now known as e0ne_14:11
*** e0ne_ is now known as e0ne14:13
*** sslypush1nko is now known as sslypushenko14:20
*** yfried__ is now known as yfried|afk14:21
*** rdekel has joined #openstack-qa14:21
*** neeti has joined #openstack-qa14:21
*** rdekel1 has joined #openstack-qa14:23
*** sabeen has joined #openstack-qa14:24
*** rdekel has quit IRC14:25
dtroyerianw: wow, dragging DevStack tests into the 2000's, thanks14:25
dtroyeris that series ready?  I'm heading off to see how bad things still are on OS/X14:26
*** mwagner_lap has quit IRC14:29
*** sabeen2 has joined #openstack-qa14:32
*** dims__ has quit IRC14:33
*** zz_jgrimm is now known as jgrimm14:34
*** neeti_ has joined #openstack-qa14:35
*** sabeen has quit IRC14:36
*** neeti has quit IRC14:37
*** rdekel1 has quit IRC14:40
*** marun has joined #openstack-qa14:41
*** neeti_ has quit IRC14:41
*** neeti has joined #openstack-qa14:41
*** mtanino has joined #openstack-qa14:41
*** neeti has quit IRC14:45
openstackrecheckopenstack-dev/devstack change: https://review.openstack.org/174147 failed gate-tempest-dsvm-postgres-full in the gate queue with an unrecognized error14:46
*** DieLove has joined #openstack-qa14:49
sdaguedtroyer: the test updates?14:50
dtroyersdague: yeah, it's still working as well as expected14:51
sdagueyep14:52
sdagueyeh, those look all reasonable to me14:52
*** DieLove has quit IRC14:52
openstackgerritAndrew Boik proposed openstack-dev/devstack: Set DevStack to dual stack by default in Kilo+  https://review.openstack.org/16085614:52
sdaguedtroyer: so, we should probably figure out what else we want in before the stable/kilo branch is cut14:53
dtroyerI wasn't sure if he was done iterating on those, looks like maybe so14:53
sdaguehttps://review.openstack.org/#/c/174370/ is the last thing I'd really like, because I want to use that from grenade and drop the custom function14:53
sdaguethe stackrc default change to add back n-crt is in the gate now, that will make people's lives happier14:54
sdaguealthough, I have to say, it's not clear to me why n-crt is a dedicated service14:54
sdagueyou have any history there?14:55
dtroyersdague: it is just to support ec2 stuff, I think the reason that people want it back is because the accrc script creates ec2 creds14:55
dtroyerI'd rather make all of that opt-in, but haven't had the time to dig into that14:56
sdagueso there are a bunch of devstack things that poke it during stack.sh14:56
dtroyerI don't think that is a big deal before stable/kilo though14:56
*** yfried|afk is now known as yfried__14:59
sdagueok15:00
*** yamahata has joined #openstack-qa15:01
sdagueyeh, maybe it's just the calls down into create_userrc15:01
dtroyerwhat about sean's dual-stack default?15:02
dtroyerhttps://review.openstack.org/#/c/160856/15:02
*** dims__ has joined #openstack-qa15:04
*** yfried__ is now known as yfried|afk15:09
*** armax has joined #openstack-qa15:22
jogoclarkb: do you have a link showing shelve working across nodes?15:22
*** dims__ has quit IRC15:22
jogoI spot checked a few cases and didn't see any that worked when moving between nodes15:23
clarkbjogo: when I first got the job working15:23
clarkbI can go find logs from that time period15:23
jogoclarkb: shelve doesn't always move between nodes though15:23
clarkbI know15:23
clarkbbut the fail rate was near 100% when it didn't work15:24
jogoif this is a regression that is a really useful piece of information15:24
clarkbI then tracked down the bug which thankfully already had a fix in the works and begged nova core to get that in15:24
jogoor maybe if this is a multihost related15:24
clarkbonce that got in the job worked fine15:24
*** yfried|afk is now known as yfried__15:26
jogoits possible we didn't notice this since this doesn't happen frequently enough15:26
jogoclarkb: well I have more news:15:28
clarkbhttp://logs.openstack.org/39/174139/6/experimental/check-tempest-dsvm-multinode-full/288b165/logs/screen-n-cpu.txt.gz I think that one shelved successfully15:28
jogoclarkb:  we have other bugs as well http://logs.openstack.org/13/173213/4/check/check-tempest-dsvm-multinode-full/9ac7736/logs/testr_results.html.gz15:28
clarkblooks like that log shows the node going down then the other shows it coming back?15:28
jogothat is from a job should have passed15:28
clarkbjogo: well did you see my comment about the dhcp failures?15:29
clarkbI think thats a nova bug too15:29
clarkbbut need to track it down with a bit more data15:29
jogoclarkb: btw https://review.openstack.org/#/c/169449/15:29
jogois still waiting for a +W15:29
clarkbI am wondering if that should go in after my ansible change15:30
clarkbalso I do not know why it wasn't approved15:31
jogoclarkb: you are right the log you sent shows TestShelveInstance-2141082895 working between nodes15:31
clarkbjogo: and that run was from yesterday15:32
jogohttp://logs.openstack.org/13/173213/4/check/check-tempest-dsvm-multinode-full/9ac7736/logs/10.208.198.2-subnode/screen-n-cpu.txt.gz?level=TRACE#_2015-04-17_10_57_46_35715:34
jogougg15:34
jogomore ebtables issues15:34
jogocburgess: ^15:34
cburgessGrrrr15:34
cburgessjogo: Did you merge that other patch we worked on?15:35
*** macjack has joined #openstack-qa15:35
cburgessjogo: Oh this is cute... thats actually libvirt now telling us it had a race.15:35
jogocburgess: I wouldn't call that cute but yeah15:36
cburgessI'm starting to think that we need to have libvirt do all ebtables work and we need to *not* touch ebtables at the nova level.15:36
jogocburgess: yeah ...15:36
cburgessOtherwise we will need to have both nova and libvirt use/support the lock file which will require sufficiently new enough versions of libvirt and ebtables.15:37
jogokashyap: ^15:37
*** nfedotov has joined #openstack-qa15:37
jogoand it gets ugly quickly15:37
jogolets move this to the nova room cburgess15:37
cburgessOr just use a sufficiently old enough version of ebtables where it just ignore the problem. :)15:37
clarkbjogo: so can we abandon that test skip?15:37
clarkband focus on fixing these bugs instead?15:37
kashyapjogo: 1 moment, on a phone call. Be back in a bit.15:38
jogoclarkb: I'll mark it as a WIP for now15:38
jogoI think we should time  box the fix for this test15:38
jogootherwise revisit it later15:38
clarkbwhat no15:39
*** Swami has joined #openstack-qa15:39
clarkbthis isn't a relatively major bug in nova just prior to release?15:39
clarkbnova will just drop nodes on the floor because raisins15:39
jogoclarkb: it is fairly big, but not clear if this is a regression or not15:41
jogoand its a fairly minor API call15:41
clarkbjogo: looking at your ebtables stuff I think that is what breaks shelve15:41
clarkbjogo: at least for the job log you linked above15:41
cburgessSo this is a byproduct of newer kernels and ebtables. If you run nova on old enough kernel/ebtables this isn't an issue.15:41
clarkbjogo: so its not a minor nova feature15:41
clarkbjogo: its booting nodes in general15:41
cburgessFor instance.. gate would never have seen this on stock precise.15:42
clarkb(also appears that it may be what breaks dhcp)15:42
clarkbcburgess: we shouldn't expect deployers to deploy nova on precise :)15:42
clarkbit only has 2 years of support left >_>15:43
cburgessclarkb: Completely agree, just stating that for regression vs not, old versions of nova would have the same problem on trusty etc.15:43
clarkbgotcha15:43
jogoclarkb: clarkb so we are talking about two separate bugs15:43
clarkbjogo: I think they are related15:43
jogo1) libvirt+ebtables15:43
sdaguedoes libvirt always call ebtables via a shell out?15:43
clarkbjogo: unless you have specific instances of shelve that fail without ebtables exploding15:43
jogo2) shelve unshelve15:43
sdaguelike... could our fix be an ebtables wrapper that does a big lock15:43
clarkbjogo: I have not seen evidence of the shelve unshelve breakage15:43
jogoclarkb: yes most cases don't explode with ebtables like this15:44
clarkbjogo: other than your link above which is ebtables15:44
*** Longgeek_ has quit IRC15:44
cburgesssdague: ebtables already supports a lock, its a matter of knowing you have a new enough version of ebtables and having a new enough version of libvirt that supports it.15:44
*** Guest43836 has quit IRC15:44
sdaguecburgess: right, but we don't know that15:45
sdagueso we can work around it on our side15:45
sdagueuntil such time as we can know that15:45
jogoclarkb: http://logs.openstack.org/88/174288/1/check/check-tempest-dsvm-multinode-full/3e3be58/console.html15:47
sdagueanyway, lunch time, later folks15:47
jogoclarkb: shelve unshelve break15:47
openstackgerritMerged openstack-dev/devstack: Revert "List all CIRROS ARCH images in image_list.sh"  https://review.openstack.org/17093415:47
clarkbjogo: http://logs.openstack.org/88/174288/1/check/check-tempest-dsvm-multinode-full/3e3be58/console.html#_2015-04-16_11_26_00_029 no thats dhcp break15:47
jogoclarkb: with no stacktrace15:47
cburgesssdague: Yeah the trick is I'm not sure about getting libvirt to use that wrapper. It would probably have supportability issues with different distros (aka I can't image redhat being cool with that).15:47
jogoclarkb: yes, dhcp breaks for shelve/unshelve15:47
clarkbjogo: its dhcp breaking, so next step is to go look in syslog logs for the requests15:47
sdaguecburgess: yeh, it could be optional15:48
jogoclarkb: agreed. Having two different conversations at the same time here :/15:48
clarkbjogo: ltes back up15:48
*** yfried__ has quit IRC15:48
clarkbjogo: dhcp breaks in gneral not just for shelve unshelve15:48
*** enikanorov__ has joined #openstack-qa15:48
jogocburgess: what about newer libvirt?15:48
clarkbjogo: if shelve unshelve breaks and dhcp is involved then that isn't likely due to shelves fault15:48
clarkbjogo: instead it is an issue iwth dhcp15:49
openstackgerritMerged openstack-dev/devstack: Remove unused IPv6 setup from DevStack neutron script  https://review.openstack.org/16187715:49
openstackgerritMerged openstack-dev/devstack: Added libxml2-dev package as a pre-requisite when installing tempest. This is required so that devstack can be installed on vanilla Ubuntu systems.  https://review.openstack.org/17113815:49
clarkbjogo: I think there is likely a race between nova setting up dhcp rules for the VM and the VM booting15:49
openstackgerritMerged openstack-dev/devstack: Assurance status of rabbitmq is running  https://review.openstack.org/16870515:49
openstackgerritMerged openstack-dev/devstack: Added ceilometer-agent-ipmi support  https://review.openstack.org/16709015:49
openstackgerritMerged openstack-dev/devstack: Switching Sahara to https in case of USE_SSL=True  https://review.openstack.org/15374115:49
jogoclarkb: its possible15:49
openstackgerritMerged openstack-dev/devstack: Added support of sahara with tls-proxy service  https://review.openstack.org/15430315:49
clarkbjogo: one way to confirm that would be to find dnsmasq logs in syslog for reuqests from that node that are rejected because dnsmasq does not know about that node15:49
cburgessjogo: So there is a version of libvirt that supports the lock, not sure of the exact version, it was added sometime after Paris when vish and I brought this up to Berrange.15:49
cburgessjogo: Gets back into the whole knowing you have new enough versions and what to do if you don15:50
cburgesser.. don't15:50
jogohttp://logs.openstack.org/88/174288/1/check/check-tempest-dsvm-multinode-full/3e3be58/logs/10.176.200.184-subnode/syslog.txt.gz15:50
*** tsekiyama has joined #openstack-qa15:51
*** enikanorov_ has quit IRC15:51
kashyapcburgess: jogo I think more than half of these issues will be gone if we're using a moderately less ancient version of libvirt.15:52
jogoclarkb: Apr 16 11:16:53 devstack-trusty-2-node-rax-iad-2194251-1687 dnsmasq-dhcp[22310]: DHCPDISCOVER(br100) fa:16:3e:fb:3e:3e no address available15:52
cburgessjogo, clarkb I don't think we can move all of this into libvirt. Some of it we probably could, but some of it related to floating IPs which could exist locally on the hypervisor (multi_host) or on a seperate node.15:52
clarkbjogo: http://logs.openstack.org/88/174288/1/check/check-tempest-dsvm-multinode-full/3e3be58/logs/10.176.200.184-subnode/syslog.txt.gz#_Apr_16_11_16_00 too15:53
cburgesskashyap: Yes we could just require a certain version of libvirt and ebtables and say if you don't have them, no locking, your mileage may vary.15:53
kashyapcburgess: The thing is day in and day out, I test from newest versions of libvirt and QEMU, and keep wishing we'd move to somewhat newer.15:53
cburgessWe could provide an "alternate" wrapper script like sdague suggested in those cases with documentation on how to deploy it if you simply can't upgrade.15:53
kashyapThis is not some legacy version of OpenStack, since this is _upstream_ OpenStack development,15:54
cburgesskashyap: No argument hear, we use much newer versions of that stuff in prod. I actually "fixed" this issues about a year ago for us.15:54
kashyapI'd expect us to be fairly ahead15:54
*** sslypushenko has quit IRC15:54
kashyapBut that horse of version capabilities has been beaten beyond death on the lists without any solid conclusions. /me stops the rant here :-)15:55
cburgessjogo, clarkb The more I look at this, the more I think this is *only* an issue when nova-network and libvirt are on the same host. So multi_host and/or devstack.15:55
jogocburgess kashyap: I want us to use a newer version of libvirt as well15:55
clarkbjogo: so I think that is the bug and syslog confirms it15:55
*** sslypushenko has joined #openstack-qa15:55
jogoclarkb: maybe15:55
kashyapjogo: I know you support that view from a recent #openstack-nova conversation15:55
jogoclarkb: I am a bit distracted15:55
kashyapThat allows to focus on _real_ bugs15:55
jogocburgess clarkb: so I have to head out to coachella now ...15:55
jogowill be back monday15:55
clarkbjogo: have fun, but my -1 still stands on that test :) this appears to be a nova bug which I can write up for better tracking now that we seem to have real info15:56
* kashyap needs to head out too, if there's a discussion on the list, I'll follow 15:56
cburgessjogo: Have fun, I'll think on this some more. I might right a patch on Sunday that tries to use the latest ebtables/libvirt to solve this. Its a bit complicated though.15:56
cburgessclarkb This is really a continution of 131662115:57
jogoclarkb: thanks, yeah I am fine with a -1 on my skip for now15:57
jogoespecially since we have more information15:57
jogoclarkb: can you spearhead the discussion about moving to a new libvirt with cburgess?15:57
clarkbcburgess: sorry I may me confusing you with the other conversation jogo and I are having15:57
cburgessclarkb: Oh sorry.15:58
cburgessMy bad.15:58
clarkbjogo: cburgess we are not going to be able to just upgrade libvirt everywhere15:58
jogoclarkb: as hunting down bugs in an anciant libvirt isn't really a win15:58
clarkbjogo: but there is no way today or tomorrow to just "use latest libvirt"15:58
cburgessSo again.. its not really a libvirt bug. It has to do with how the kernel changed to handle certain calls for net filter related things.15:58
jogoclarkb: what about less old libvirt but not 'ewest'15:58
clarkbI think even russellb found that it was a bit of trouble15:58
clarkbjogo: even that isn't really possible15:58
clarkbwe don't have libvirt packages for f2X, centos7, precise, trusty15:59
jogoanyway I have to head out, I am interested to know what are options actually are15:59
* jogo gets ready for coachella15:59
clarkband until someone volunteers to do that work it is problematic15:59
cburgessWell... so... 2 possible solution here then..15:59
cburgessOr actually.. we do both.15:59
cburgess1) We fix it properly in the code assuming new versions15:59
cburgess2) We also provide a work-around script like sdague mentioned that we can enable as a work around and we configure the gate to use that.16:00
cburgessbut fundamentally its a race between nova and libvirt. We can't fix that unless we get libvirt to do *something*, even if that is call our wrapper script.16:00
*** armax has quit IRC16:01
clarkbcburgess: how new of libvirt are we talking? and is it newer than what is in f21, centos7 and ubuntu trusty?16:02
*** safchain has quit IRC16:02
cburgessclarkb: Fairly now.. http://libvirt.org/git/?p=libvirt.git;a=commit;h=dc33e6e4a5a5d429198b2c63ff6b63729353e2cf, looks like 1.2.11 according to a comment from mriedem in the bug.16:04
cburgesser fairly new16:04
clarkbya so that isn't going to be sanely possible until we have people building packages for all the things16:05
clarkbmordred: ^16:05
clarkbrussellb did have a job at one point that built libvirt from source on f20 iirc and even that had trouble keeping up.16:05
cburgessOK well, like I said, we can basically provide 2 fixes in the code. One that does the right thing with the right version and one that allows you to use a wrapper script. The wrapper script will require some documentation though since you will have to make system level changes to get libvirt to call it too.16:06
* mordred reads scrollback16:07
cburgessmordred: This is just the same old we need newer libvirt because.. reasons.16:07
mordredyah16:07
*** dkranz has quit IRC16:07
cburgessIn this case the reason is a race we can't really work around any other way.. but still its reasons.16:07
mordredand we need the general ability to be able to express such a need16:08
mordredyup16:08
clarkbhowever16:08
*** yamahata has quit IRC16:08
clarkbit is also pretty important to recognize that if your users can't use centos7 and ubuntu trusty libvirt they aren't likely to be happy users16:08
cburgessNot to down play that this a fairly ugly race and bug.16:08
mordredclarkb: I'm not sure I agree16:08
clarkbthis is a different situation than "we would like to be able to test nova's functionality with newer libvirt features"16:09
mordredclarkb: I think if it's important to a user to use trusty libvirt, it'll also be important to that user to use trusty nova16:09
clarkbthis is "well we wrote a bug around libvirt that we know about BUT everyone just upgrade libvirt"16:09
clarkbmordred: yes but trusty nova is broken too16:09
mordredif the user is willing to use a nova that was not shipped with trusty16:09
clarkbmordred: and so is cnetos7 nov16:09
cburgessEveryone is broken.16:09
mordredthe user needs to be willing to use a newer libvirt too16:09
mordredbecause the two are pretty tied together16:09
mordredand being willing to do one and not the other is bananas16:09
clarkbmordred: no one is saying do one and not the other16:10
mordred16:08:35           clarkb | it is also pretty important to recognize that if your users can't use centos7 and ubuntu trusty libvirt they aren't likely to be happy users16:10
cburgessAs ugly and disgusting as it sounds to me I think sdague wrapper script might be the only solution for those who can't upgrade.16:10
clarkbmordred: I am saying there is a pretty big flaw in all novas16:10
clarkbmordred: regardless of libvirt version16:10
mordredok16:10
clarkbmordred: yes with centos7 and trusty nova16:10
*** pcaruana has quit IRC16:10
cburgessclarkb: Assuming a new enough kernel too. The flaw doesn't exist before a certain kernel rev (can't recall off the top of my head, maybe 3.13).16:10
clarkbok so maybe just all trusty users then16:11
cburgessPoint being.. its a bad bug. I agree.16:11
cburgessSo I think we have to provide the proper fix, and a work around fix. The gate will have to use the work around for now as well *most* distro customers.16:11
*** sabeen2 has quit IRC16:11
*** e0ne has quit IRC16:12
clarkbmordred: but the distinction is important because if the only place our software works is on our test bed then we haven't done a good job16:12
clarkbcburgess: and ya it should be straightforward to have devstack install the workaround if it detect it is necessary16:13
clarkbcburgess: and that will get it into the gate while we sort out ability to test latest/newer/different libvirt16:13
*** jlanoux has quit IRC16:14
mordredclarkb: totally agree16:16
mordredclarkb: which is why I think we can't just do a one-off solution16:16
mordredclarkb: but need to have a fully fleshed out and considered mechanism16:16
*** macjack has quit IRC16:17
*** marcusvrn has joined #openstack-qa16:17
*** luqas has quit IRC16:18
*** salv-orl_ has quit IRC16:19
*** mriedem1 has joined #openstack-qa16:19
cburgessclarkb: OK I'll talk to jogo on monday and if I have time today/this weekend I'll work on some type of fix. I need to even think about what the wrapper needs to do. Basically use oslo.concurrency to do the locking and exec of ebtables.16:20
*** subscope has joined #openstack-qa16:20
*** mriedem has quit IRC16:21
*** marcusvrn1 has joined #openstack-qa16:22
*** marcusvrn2 has joined #openstack-qa16:22
*** marcusvrn has quit IRC16:23
*** marcusvrn3 has joined #openstack-qa16:23
cburgessLooks like we can write the wrapper to mimic ebtables with --concurrent such that new enough libvirts will detect it and do the right thing. We can then have nova detect if it supports --concurrent as well and if it doesn't log that a user will need to deploy the wrapper work around and provide a link to the docs about it.16:24
*** salv-orlando has joined #openstack-qa16:24
*** dims__ has joined #openstack-qa16:24
clarkbcburgess: jogo filed https://bugs.launchpad.net/nova/+bug/1445569 for the dhcp thing16:25
openstackLaunchpad bug 1445569 in OpenStack Compute (nova) "No dhcp lease after shelve unshelve" [Undecided,New]16:25
clarkbI haven't been able to say if it is specific to shelve unshelve yet, but hopefully provided enough info there for someone with nova familiarity to figure it out quickly16:25
*** tkammer has quit IRC16:26
cburgessclarkb: OK I haven't looked at that one. I16:26
*** marcusvrn1 has quit IRC16:26
cburgessI can try and poke at it if I have time today, short day for me, have to meet a repair man to get my oven fixed.16:26
*** marcusvrn2 has quit IRC16:26
*** patrickeast has joined #openstack-qa16:27
cburgessclarkb: Taking a quick look the proper calls seem to be in the code to update dnsmasq with the network info. Didn't do a deep trace.16:29
*** tkammer has joined #openstack-qa16:30
*** yamahata has joined #openstack-qa16:31
cburgessclarkb: Does that shelve/unshelve fail everytime?16:33
clarkbcburgess: no16:33
*** Nic has joined #openstack-qa16:33
cburgessIs shelve offload called as part of the test?16:34
*** nfedotov has quit IRC16:37
*** dkranz has joined #openstack-qa16:38
clarkbcburgess: https://git.openstack.org/cgit/openstack/tempest/tree/tempest/scenario/test_shelve_instance.py#n57 appears to yes16:40
cburgessclarkb: k... this is weird.. trying to understand this code flow.16:41
*** armax has joined #openstack-qa16:46
*** dims__ is now known as dimsum__16:50
*** ajmiller__ is now known as ajmiller16:53
*** jlibosva has quit IRC16:55
*** Nic has quit IRC16:56
*** armax has quit IRC16:58
*** ihrachyshka has quit IRC17:00
*** armax has joined #openstack-qa17:03
*** dimtruck is now known as zz_dimtruck17:05
*** salv-orlando has quit IRC17:06
*** suneelb has joined #openstack-qa17:09
*** catohornet has quit IRC17:09
*** harlowja_away is now known as harlowja17:17
*** luqas has joined #openstack-qa17:19
*** Nic has joined #openstack-qa17:20
*** luqas has quit IRC17:24
*** salv-orlando has joined #openstack-qa17:29
*** catohornet has joined #openstack-qa17:35
openstackgerritMerged openstack-dev/devstack: Remove keystone public/admin_endpoint options  https://review.openstack.org/16877017:36
*** e0ne has joined #openstack-qa17:37
*** mriedem1 is now known as mriedem17:38
openstackgerritMerged openstack-dev/devstack: Fixing n-crt removal from stackrc  https://review.openstack.org/17349317:42
*** jordanP has quit IRC17:43
*** gfidente is now known as gfidente|afk17:45
*** e0ne is now known as e0ne_17:45
*** achanda has joined #openstack-qa17:48
*** e0ne_ is now known as e0ne17:52
*** e0ne is now known as e0ne_17:57
*** vponomaryov has quit IRC18:01
*** armax has quit IRC18:05
*** aix has quit IRC18:06
openstackrecheckopenstack/cinder change: https://review.openstack.org/174783 failed gate-tempest-dsvm-full in the gate queue with an unrecognized error18:11
mtreinishwow, ddt makes some ugly test_ids18:17
*** Sukhdev has joined #openstack-qa18:19
*** armax has joined #openstack-qa18:20
*** mlavalle has joined #openstack-qa18:24
*** leakypipes has quit IRC18:25
*** rushiagr is now known as rushiagr_away18:26
*** dustins_ has joined #openstack-qa18:26
*** annashen has joined #openstack-qa18:27
*** johnthetubaguy is now known as zz_johnthetubagu18:28
*** andreykurilin__ has joined #openstack-qa18:29
*** dustins has quit IRC18:29
*** e0ne_ is now known as e0ne18:30
*** cdent has quit IRC18:36
*** tosky has quit IRC18:42
*** achanda has quit IRC18:46
*** annashen has quit IRC18:47
mtreinishdkranz: if you get a sec, I've got a bunch of doc patches up: https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:tempest-config-guide,n,z18:51
mtreinishsome have -1s but I'd like to just update in a batch18:52
mtreinishso if you've got more comments :)18:52
mtreinishhogepodge: ^^^ you too :)18:54
*** achanda has joined #openstack-qa18:56
*** salv-orlando has quit IRC18:57
*** e0ne has quit IRC18:58
*** marcusvrn3 has quit IRC19:00
*** marun has quit IRC19:03
dkranzmtreinish: on it in a few19:13
mtreinishdkranz: cool thanks, no rush19:13
sdaguemtreinish: there are 2 grenade patches you could approve before the weekend - https://review.openstack.org/#/c/174302/3 :)19:18
sdagueok, running away, arwen is waking up19:18
mtreinishsdague: done19:19
*** bnemec is now known as beekneemech19:22
pc_mmtreinish: hi19:22
mtreinishpc_m: hi19:23
pc_mmtreinish: I'm hitting an issue with the VPN functional job that I'm hoping you may have some thoughts on.19:24
*** suneelb has quit IRC19:24
pc_mmtreinish: Are the gate-hook.sh scripts supposed to run in strict bash mode?19:24
mtreinishpc_m: you mean the strict posix mode?19:25
mtreinishor something else19:25
pc_mmtreinish: The equivalent of bash with set -u.19:25
pc_mmtreinish: It seems like the VPN gate-hook.sh script is running in strict mode, and the neutron one is not. I cannot figure out where this could be getting set as the two repos have very similar files for everything.19:27
*** dustins_ has quit IRC19:27
*** ajo has quit IRC19:28
*** ajo has joined #openstack-qa19:28
*** ajo has quit IRC19:28
mtreinishpc_m: well looking at the job definition it's set to call bash -xe19:30
mtreinishpc_m: http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/neutron-vpnaas.yaml#n2819:30
openstackgerritMerged openstack-dev/devstack: Add "passed" and "failed" functions  https://review.openstack.org/17468219:30
mtreinishwhich is what gets passed to: http://git.openstack.org/cgit/openstack-infra/devstack-gate/tree/functions.sh#n4619:30
pc_mmtreinish: Yeah, same as neutron.yaml.19:30
*** yamahata has quit IRC19:31
mtreinishso unless something inside the scripts (or something it's calling is calling set -u) I'm not sure19:31
openstackgerritMerged openstack-dev/devstack: Use unittest help in test_ip.sh  https://review.openstack.org/17468319:31
openstackgerritMerged openstack-dev/devstack: Fix return of test_ini_config & test_meta_config  https://review.openstack.org/17468419:31
openstackgerritMerged openstack-dev/devstack: Move function.sh to test_functions.sh  https://review.openstack.org/17468519:31
openstackgerritMerged openstack-dev/devstack: Fix die_if_not_set tests  https://review.openstack.org/17468619:32
openstackgerritMerged openstack-dev/devstack: Convert test_functions.sh to use unittest helpers  https://review.openstack.org/17468719:32
openstackgerritMerged openstack-dev/devstack: Remove old comment in run_tests.sh  https://review.openstack.org/17468819:32
openstackgerritMerged openstack-dev/devstack: Create config file in merge_config_file  https://review.openstack.org/17467819:32
pc_mmtreinish: The VPN repo cloned and modified the neutron.yaml, and ditto for the gate-hook.sh. So they seem to be the same, but when I run vpn functional job, it is in -u mode.19:32
pc_mmtreinish: So I don't know how it is getting turned on, in the vpn case.19:34
pc_mmtreinish: Is it possible that some other set option would automatically set -u?19:34
mtreinishpc_m: I assume it is one of these sources: http://git.openstack.org/cgit/openstack/neutron-vpnaas/tree/neutron_vpnaas/tests/contrib/gate_hook.sh#n2519:35
mtreinishbut I didn't trace through it all, but I could see set -u being called by sourcing a bunch of things from devstack19:35
pc_mmtreinish: In the current VPN tree, it's not running with -u (as everything works fine). The gate-hook stacks DevStack via the devstack-vm-gate.sh script19:36
pc_mmtreinish: In my patchset, I'm trying to do what Neutron has done, and only configure devstack, but not actually stack.19:37
pc_mmtreinish: Like neutron, it ends up calling some devstack scripts, that assume you are NOT running in -u mode. For neutron, it works, for VPN, it fails with an unbound variable error.19:38
mtreinishpc_m: so it's your patch that's failing do you have a link?19:39
pc_mmtreinish: I guess I could unset -u, but wanted to understand why neutron works.19:39
pc_mmtreinish: yes... https://review.openstack.org/#/c/168115/1519:39
pc_mHere's the failing log... http://logs.openstack.org/15/168115/17/check/check-neutron-vpnaas-dsvm-functional/261ea6d/console.html19:40
*** subscope has quit IRC19:40
pc_mmtreinish: It calls a method _get_package_list, indirectly, and that method expects a $1 arg, and then checks to see if it is empty. Being in -u mode, the job fails.19:41
mtreinishpc_m: http://logs.openstack.org/15/168115/17/check/check-neutron-vpnaas-dsvm-functional/261ea6d/console.html#_2015-04-17_11_33_47_39419:41
pc_mmtreinish: If you look at any Neutron commit, the same invocation of the script passes.19:41
mtreinishthis source: source /opt/stack/new/neutron/tools/deploy_rootwrap.sh /opt/stack/new/neutron /opt/stack/new/neutron-vpnaas/.tox/dsvm-functional/etc /opt/stack/new/neutron-vpnaas/.tox/dsvm-functional/bin19:41
mtreinishis calling set -eu19:41
pc_mmtreinish: Ah! I was looking for -u in the log.19:42
pc_mmtreinish: Thanks. I was going crazy.19:42
mtreinishsure, np19:42
*** ibiris is now known as ibiris_away19:45
mtreinishdpaterson: hey, if you get a chance could you post up your cli work? I think I'll have some time next week to play with it19:48
mtreinisheven if doesn't come close to working it'd be good to have the reviews up19:48
*** annashen has joined #openstack-qa19:48
*** madhu_ak has joined #openstack-qa19:49
*** salv-orlando has joined #openstack-qa19:49
dpatersonmtreinish: yes I will this weekend, sorry for delay, up to eyeballs in other stuff.19:50
*** achanda has quit IRC19:50
mtreinishdpaterson: sure, no worries. I just wanted to make sure it didn't fall through the cracks :)19:50
dpatersonmtreinish: Sure, and thanks19:51
*** annashen has quit IRC19:53
*** patrickeast has quit IRC19:53
*** tkammer has quit IRC19:55
madhu_akHello19:56
madhu_ak, I have created tempest tests in neutron-lbaas tree and it perfectly working as designed for admin and non-admin users..the resources are indeed setup and cleaned up against devstack instances. For admin tests, when i run the tests for many times and upon completion, I could see bunch of tenant_id's when I randomly check 'keystone tenant-list'.19:56
madhu_akI am not sure where is the best place to write a piece of code for deleting the tenant_ids..19:57
madhu_akupon researching, I found this: https://github.com/stackforge/ospurge, I am not familiar with it though. Do we use ospurge for other projects?19:57
*** annashen has joined #openstack-qa20:00
madhu_akcan anybody help?20:02
*** gfidente|afk has quit IRC20:09
*** artom has quit IRC20:11
mtreinishmadhu_ak: if the resources aren't being cleaned up then you did something wrong in your setupclass/teardownclass pieces20:11
mtreinishbecause it should always be tearing down whatever it creates20:12
mtreinishyou don't want to have users need to rely on ospurge or tempest's cleanup: http://git.openstack.org/cgit/openstack/tempest/tree/tempest/cmd/cleanup.py to cleanup after your tests20:12
*** jogo has quit IRC20:14
madhu_akso, having resource_cleanup method will clear the tenant_ids as well?20:14
*** dkranz has quit IRC20:14
mtreinishmadhu_ak: you should look at the tempest code around using tenant isolation in test.py for setupclass and teardownclass and look at what existing tempest tests are doing20:15
mtreinishyou'll see how to invoke things to ensure they get cleaned up20:15
madhu_akoh, thanks for that info..will look into it.. I think calling tearDownClass method will cleanup the resources20:17
*** Sukhdev has quit IRC20:19
madhu_akbtw, do we have sample code for tearDownClass or resource clean_up for admin test cases, maybe are we using like this in neutron/compute/or other projects?20:20
*** openstackgerrit has quit IRC20:22
*** openstackgerrit has joined #openstack-qa20:22
openstackgerritMerged openstack-dev/grenade: Comment function files  https://review.openstack.org/17430220:23
openstackgerritMerged openstack-dev/grenade: ensure that grenade db is fresh on each run  https://review.openstack.org/17430320:23
openstackgerritMerged openstack-dev/grenade: update README.rst to reflect current reality  https://review.openstack.org/17381720:23
*** marun has joined #openstack-qa20:25
*** achanda has joined #openstack-qa20:27
*** Sukhdev has joined #openstack-qa20:28
*** annashen has quit IRC20:28
*** andreykurilin___ has joined #openstack-qa20:28
*** andreykurilin__ has quit IRC20:29
*** annashen has joined #openstack-qa20:32
*** mlavalle has quit IRC20:34
*** annashen_ has joined #openstack-qa20:44
*** annashen has quit IRC20:44
*** luqas has joined #openstack-qa20:57
madhu_akmtreinish: thanks for that, I just used resource_cleanup method, its deleting the tenant_ids :-)21:01
*** armax has quit IRC21:01
*** luqas has quit IRC21:01
openstackgerritSwaminathan Vasudevan proposed openstack-infra/devstack-gate: DVR Multinode test patch to separate ctlr and compute  https://review.openstack.org/17505621:05
-openstackstatus- NOTICE: Gerrit will be unavailable between 22:00 and 23:59 UTC for project renames and a database update.21:05
*** dpaterson has quit IRC21:06
*** mriedem has quit IRC21:07
*** BadCub is now known as Party_Bear21:17
*** pc_m has quit IRC21:18
*** shardy has quit IRC21:20
*** openstackgerrit has quit IRC21:23
*** openstackgerrit has joined #openstack-qa21:23
openstackrecheckopenstack/nova change: https://review.openstack.org/174524 failed gate-tempest-dsvm-postgres-full in the gate queue with an unrecognized error21:26
*** dkranz has joined #openstack-qa21:26
*** erlon has quit IRC21:31
*** annashen_ has quit IRC21:36
*** annashen has joined #openstack-qa21:37
*** ArxCruz has joined #openstack-qa21:41
*** bknudson has quit IRC21:44
*** armax has joined #openstack-qa21:47
dkranzmtreinish: I'm reviewing your doc patches (thank you!) and am wondering how much grammatical nitpickiness is appropriate. Do you want a -1 and correction for every little thing?21:48
*** jamielennox|away is now known as jamielennox21:48
dkranzmtreinish: I'm ok either way21:49
mtreinishdkranz: I'm gonna have to rebase the whole stack probably so -1 away so the stat trackers will pick it up :)21:52
*** mlavalle has joined #openstack-qa21:58
*** tkammer has joined #openstack-qa22:02
*** jecarey has quit IRC22:04
*** Swami has quit IRC22:04
-openstackstatus- NOTICE: Gerrit is unavailable until 23:59 UTC for project renames and a database update.22:05
-openstackstatus- NOTICE: Gerrit is unavailable until 23:59 UTC for project renames and a database update.22:08
*** ChanServ changes topic to "Gerrit is unavailable until 23:59 UTC for project renames and a database update."22:08
dkranzmtreinish: ok. I ran out of time and gerrit is down,  but I will just pull the last one and fix the issues. Gerrit is not a very good UI for marking up text while editing :)22:13
*** andreykurilin___ has quit IRC22:13
*** annashen has quit IRC22:15
*** annashen has joined #openstack-qa22:16
*** jgrimm is now known as zz_jgrimm22:20
*** Sukhdev has quit IRC22:28
*** dboik_ has quit IRC22:41
*** davideagnello has joined #openstack-qa22:41
*** dboik has joined #openstack-qa22:41
*** annashen has quit IRC22:41
*** ArxCruz has quit IRC22:42
*** Sukhdev has joined #openstack-qa22:51
*** achanda has quit IRC22:54
*** annashen has joined #openstack-qa22:54
*** aysyd has quit IRC22:56
*** achanda has joined #openstack-qa22:56
*** dboik has quit IRC23:02
*** ChanServ changes topic to "All Things QA..."23:05
-openstackstatus- NOTICE: Gerrit is available again.23:05
*** jhenner has quit IRC23:09
*** patrickeast has joined #openstack-qa23:10
*** achanda has quit IRC23:11
*** annashen has quit IRC23:15
*** davideagnello has quit IRC23:15
*** achanda has joined #openstack-qa23:15
*** mlavalle has quit IRC23:20
*** Swami has joined #openstack-qa23:28
*** dboik has joined #openstack-qa23:29
*** jamielennox is now known as jamielennox|away23:30
*** flaper87 has quit IRC23:38
*** ajmiller has quit IRC23:45
*** patrickeast has quit IRC23:50
*** dboik has quit IRC23:57
*** weshay has quit IRC23:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!