Wednesday, 2019-01-23

*** tosky has quit IRC00:01
*** mriedem has quit IRC00:02
*** vegasq has quit IRC00:21
*** artom has joined #openstack-qa00:23
*** spsurya has quit IRC00:26
openstackgerritMerged openstack-dev/devstack stable/rocky: Restrict database access to nova-compute in singleconductor mode  https://review.openstack.org/63209501:45
*** hongbin has joined #openstack-qa02:25
openstackgerritMerged openstack/tempest master: Update v3 oauth consumers tests to work w/ pre-prov  https://review.openstack.org/62865902:36
*** ekcs has quit IRC02:44
*** ramishra has joined #openstack-qa03:13
*** rh-jelabarre has quit IRC03:52
*** hongbin has quit IRC04:18
*** whoami-rajat has joined #openstack-qa04:35
*** sridharg has joined #openstack-qa05:11
*** dave-mccowan has quit IRC05:22
*** spsurya has joined #openstack-qa06:08
*** agurenko has joined #openstack-qa06:34
*** ccamposr has joined #openstack-qa06:53
*** ccamposr__ has joined #openstack-qa06:55
*** ccamposr has quit IRC06:58
*** Luzi has joined #openstack-qa06:59
*** aojea has joined #openstack-qa07:03
*** gcheresh has joined #openstack-qa07:08
*** slaweq has joined #openstack-qa07:18
*** aojea has quit IRC07:28
*** aojea has joined #openstack-qa07:29
*** aojea has quit IRC07:31
*** kopecmartin|off is now known as kopecmartin07:47
*** gfidente has joined #openstack-qa07:47
*** ramishra has quit IRC07:50
*** ramishra has joined #openstack-qa07:56
*** e0ne has joined #openstack-qa07:57
openstackgerritJuan Antonio Osorio Robles proposed openstack-dev/devstack master: Add keystone internal URL in bootstrap  https://review.openstack.org/63263508:00
*** rcernin has quit IRC08:04
*** jpena|off is now known as jpena08:14
*** gkadam has joined #openstack-qa08:22
*** rpittau has joined #openstack-qa08:28
*** shardy has joined #openstack-qa08:28
*** e0ne has quit IRC08:33
openstackgerritlijunjie proposed openstack/openstack-health master: Fix the misspelling of "register"  https://review.openstack.org/63264808:39
*** markvoelker has joined #openstack-qa08:47
*** tosky has joined #openstack-qa08:50
*** rpittau has quit IRC08:51
*** rpittau has joined #openstack-qa08:52
*** markvoelker has quit IRC08:53
openstackgerritlijunjie proposed openstack/stackviz master: Fix the misspelling of "register"  https://review.openstack.org/63265308:56
*** rpittau has quit IRC09:07
*** rpittau has joined #openstack-qa09:08
*** rpittau has quit IRC09:22
*** rpittau has joined #openstack-qa09:22
*** dtantsur|afk is now known as dtantsur09:23
*** agurenko has quit IRC09:39
*** markvoelker has joined #openstack-qa09:49
*** rpittau has quit IRC09:52
*** rpittau has joined #openstack-qa09:53
*** aojea has joined #openstack-qa09:59
*** rpittau has quit IRC10:07
*** rpittau has joined #openstack-qa10:07
*** jpena is now known as jpena|brb10:08
*** yamamoto has joined #openstack-qa10:16
*** markvoelker has quit IRC10:22
*** rpittau has quit IRC10:22
*** rpittau has joined #openstack-qa10:22
*** pcaruana has joined #openstack-qa10:37
*** rpittau has quit IRC10:37
*** rpittau has joined #openstack-qa10:40
slaweqgmann: hi10:44
slaweqgmann: may I have quick question about how tests are run?10:45
slaweqgmann: is there any possibility to mark somehow two tests from same class that they need to be run separately?10:45
gmannslaweq: hi10:47
gmannslaweq: run separately you mean?10:47
slaweqI have 2 functional tests in neutron10:47
slaweqand now 2 different workers can run them in parallel10:47
slaweqwhich sometimes can leads to failure of one of them10:48
slaweqso I would like to run them alway one by one10:48
slaweqlike with some lock or something like that10:48
gmannslaweq: i see, so you can put those in same class. so tests within class is always run serially.10:49
gmanni do not think they can run in parallel10:49
slaweqgmann: those tests are in same class10:50
gmannslaweq: if you want some isolation or sync between them you can use lockutils.LockFixture also10:50
gmannexample-   https://github.com/openstack/tempest/blob/6d4657a949c34ffd153e43d5d55bdfc0202a0a0a/tempest/api/compute/admin/test_aggregates.py#L15210:51
slaweqgmann: ok, I though that there is maybe something specific for tests but if not, I will use this lockutils, thx10:51
gmannyeah, we use that for quota, aggregate related tests where one test run can effect other test run at same time10:52
*** rpittau has quit IRC10:52
*** rpittau has joined #openstack-qa10:53
*** andreaf has quit IRC10:54
*** andreaf has joined #openstack-qa10:56
*** jpena|brb is now known as jpena11:07
*** rpittau has quit IRC11:08
*** rpittau has joined #openstack-qa11:11
*** markvoelker has joined #openstack-qa11:20
*** e0ne has joined #openstack-qa11:26
*** yamamoto has quit IRC11:50
*** yamamoto has joined #openstack-qa11:52
*** markvoelker has quit IRC11:52
*** jpena is now known as jpena|lunch12:33
*** yamamoto has quit IRC12:40
*** rh-jelabarre has joined #openstack-qa12:45
*** markvoelker has joined #openstack-qa12:49
*** artom has quit IRC13:10
*** yamamoto has joined #openstack-qa13:12
*** yamamoto has quit IRC13:22
*** markvoelker has quit IRC13:22
*** markvoelker has joined #openstack-qa13:23
*** markvoelker has quit IRC13:30
openstackgerritSergey Vilgelm proposed openstack/patrole master: Fix OverPermission exception for keystone tests  https://review.openstack.org/62902413:31
*** yamamoto has joined #openstack-qa13:31
*** yamamoto has quit IRC13:31
*** jpena|lunch is now known as jpena13:34
*** dave-mccowan has joined #openstack-qa13:43
*** yamamoto has joined #openstack-qa13:46
*** yamamoto has quit IRC13:50
*** mriedem has joined #openstack-qa13:56
*** yamamoto has joined #openstack-qa14:02
*** yamamoto has quit IRC14:06
openstackgerritMerged openstack-dev/devstack stable/queens: Restrict database access to nova-compute in singleconductor mode  https://review.openstack.org/63210014:06
*** mriedem has quit IRC14:10
*** awaugama has quit IRC14:12
*** artom has joined #openstack-qa14:21
*** artom has quit IRC14:37
openstackgerritMichał Dulko proposed openstack-dev/devstack master: Add option to place etcd data dir on RAM disk  https://review.openstack.org/62688514:48
openstackgerritMichał Dulko proposed openstack-dev/devstack master: Add option to place etcd data dir on RAM disk  https://review.openstack.org/62688514:49
*** awaugama has joined #openstack-qa14:51
*** ramishra has quit IRC15:10
openstackgerritSergey Vilgelm proposed openstack/patrole master: Fix OverPermission exception for keystone tests  https://review.openstack.org/62902415:12
*** aojea_ has joined #openstack-qa15:12
openstackgerritSergey Vilgelm proposed openstack/patrole master: Fix OverPermission exception for keystone tests  https://review.openstack.org/62902415:14
*** vegasq has joined #openstack-qa15:18
openstackgerritLajos Katona proposed openstack/tempest master: Add placement client for basic GET operations  https://review.openstack.org/62231615:22
openstackgerritLajos Katona proposed openstack/tempest master: Add QoS client for basic operations  https://review.openstack.org/62914215:22
openstackgerritLajos Katona proposed openstack/tempest master: Minimum bandwidth allocation with placement  https://review.openstack.org/62925315:22
*** vegasq_ has joined #openstack-qa15:24
*** vegasq has quit IRC15:25
*** Luzi has quit IRC15:30
*** e0ne has quit IRC15:30
*** aojea_ has quit IRC15:31
*** e0ne has joined #openstack-qa15:33
*** e0ne has quit IRC15:33
*** e0ne has joined #openstack-qa15:33
*** artom has joined #openstack-qa16:03
*** gcheresh has quit IRC16:21
*** ccamposr__ has quit IRC16:41
*** sshnaidm is now known as sshnaidm|afk16:46
*** e0ne has quit IRC16:55
openstackgerritSergey Vilgelm proposed openstack/patrole master: Fix OverPermission exception for keystone tests  https://review.openstack.org/62902416:56
*** rpittau has quit IRC16:58
*** aojea has quit IRC17:01
*** gfidente is now known as gfidente|afk17:17
*** gkadam has quit IRC17:29
*** sridharg has quit IRC17:40
*** dtantsur is now known as dtantsur|afk17:55
*** kopecmartin is now known as kopecmartin|off18:01
*** shardy has quit IRC18:21
*** ekcs has joined #openstack-qa18:22
*** aojea has joined #openstack-qa18:27
*** aojea has quit IRC18:29
*** ssbarnea|rover has quit IRC18:29
*** ssbarnea|bkp2 has joined #openstack-qa18:29
*** aojea has joined #openstack-qa18:30
*** yamamoto has joined #openstack-qa18:42
*** jpena is now known as jpena|off18:44
*** yamamoto has quit IRC18:47
*** mugsie has quit IRC18:55
*** mugsie has joined #openstack-qa18:57
openstackgerritSergey Vilgelm proposed openstack/patrole master: Fix OverPermission exception for keystone tests  https://review.openstack.org/62902418:59
openstackgerritAdam Spiers proposed openstack-dev/devstack master: Fix distro detection for SUSE Linux Enterprise  https://review.openstack.org/63279219:01
*** e0ne has joined #openstack-qa19:10
openstackgerritAdam Spiers proposed openstack-dev/devstack master: Fix distro detection for SUSE Linux Enterprise  https://review.openstack.org/63279219:12
*** e0ne has quit IRC19:32
*** e0ne has joined #openstack-qa19:40
*** gyee has joined #openstack-qa19:48
gyeewhat's the magic option to tell tempest to preserve the server after the test run?19:50
*** spsurya has quit IRC19:55
*** aojea has quit IRC20:03
*** aojea has joined #openstack-qa20:04
*** e0ne has quit IRC20:18
*** mriedem has joined #openstack-qa20:56
*** awaugama has quit IRC20:58
*** aojea_ has joined #openstack-qa21:27
*** aojea has quit IRC21:28
*** whoami-rajat has quit IRC21:45
mtreinishgmann, slaweq: tests in the same class are run serially and do not run in parallel (unless you go outside of tempest/stestr) we specifically set tempest/stestr to group tests in a class together to run serailly21:48
mtreinishthe unit of parallelization is classes not individual test methods21:48
openstackgerritSergey Vilgelm proposed openstack/patrole master: Fix OverPermission exception for keystone tests  https://review.openstack.org/62902421:49
mtreinishthe locking fixture should only really be necessary or used for global object types which can effect other tests even outside of tenant/project isolation (like aggregates in that example gmann linked)21:50
slaweqmtreinish: yes, now I know it :) I had different issue and it wasn't related to tests running in parallel21:51
slaweqsorry for noise :)21:51
slaweqmtreinish: may I ask You one more question about stestr?21:52
mtreinishsure21:52
slaweqmtreinish: can You check my mail http://lists.openstack.org/pipermail/openstack-discuss/2019-January/001904.html if You will have some time?21:52
slaweqI really need some help with it because I can't move forward with this job and python 3 :/21:53
slaweqit's not urgent but would be great if You maybe could take a look21:53
mtreinishthat's a known issue with stestr and python3 caused by excessively large attachments in the subunit stream21:54
mtreinishthe only way to workaround it is, as you mentioned in the email, to decrease the size of the attachments from the test results (stdout, stderr, and pylogging)21:55
mtreinishsmcginnis, masayukig, and I spent hours trying to find the root cause of the bug when cinder hit the issue, but came up empty21:56
slaweqmtreinish: so should we avoid every stacktrace like http://logs.openstack.org/83/577383/28/check/neutron-functional/649ba74/job-output.txt.gz#_2019-01-18_00_35_05_449643 from tests?21:59
mtreinishslaweq: something that would be helpful to try and debug is if you could run the tests in parallel (like in the gate), serially, and with subunit.run directly so we can easily locally reproduce the subunit parsing without having to set up a test run22:00
mtreinishslaweq: yeah removing stacktraces like that will help22:00
mtreinishalthough based on that log file I expect the pythonlogging in the attachments is actually what's triggering the issue22:01
mtreinishbecause the stack traces are pretty small22:01
slaweqmtreinish: are You saying that issue is caused by logs which are stored in files here: http://logs.openstack.org/83/577383/28/check/neutron-functional/649ba74/logs/dsvm-functional-logs/ ?22:02
slaweqshould we not log anything?22:02
*** rcernin has joined #openstack-qa22:03
slaweqmtreinish: I will try to run tests locally in such 3 ways22:04
mtreinishslaweq: no the pythonlogging attachment in the stream: http://logs.openstack.org/83/577383/28/check/neutron-functional/649ba74/job-output.txt.gz#_2019-01-18_00_37_25_28015322:07
mtreinishthat is much larger than the stack trace attachments (which are written to an attachment named stderr)22:07
mtreinishso if I was going to guess the part of the stream which is causing the bug in stestr w/ py3 to be triggered it's those types of attachments22:07
slaweqmtreinish: ok, thx for tip22:08
slaweqI will try to disable this logging and see then22:08
*** dave-mccowan has quit IRC22:25
*** raissa has joined #openstack-qa22:26
*** gfidente|afk has quit IRC22:50
*** yamamoto has joined #openstack-qa23:01
*** raissa has quit IRC23:06
*** aojea_ has quit IRC23:13
*** tosky has quit IRC23:43
*** mriedem has quit IRC23:48

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!