Monday, 2019-01-14

*** dave-mccowan has joined #openstack-barbican00:09
*** dave-mccowan has quit IRC00:13
*** remi_ness has quit IRC03:01
*** remi_ness has joined #openstack-barbican03:15
*** whoami-rajat has joined #openstack-barbican03:30
*** remi_ness has quit IRC06:51
*** Luzi has joined #openstack-barbican06:53
*** pcaruana has joined #openstack-barbican07:11
*** velizarx has joined #openstack-barbican08:07
*** moguimar has joined #openstack-barbican08:08
*** ssbarnea|rover has joined #openstack-barbican08:54
*** livelace has joined #openstack-barbican10:54
*** raildo has joined #openstack-barbican11:53
*** salmankhan has joined #openstack-barbican12:31
*** salmankhan1 has joined #openstack-barbican13:07
*** salmankhan has quit IRC13:08
*** salmankhan1 is now known as salmankhan13:08
*** velizarx has quit IRC13:09
*** dave-mccowan has joined #openstack-barbican13:14
*** velizarx has joined #openstack-barbican13:21
*** velizarx has quit IRC14:02
*** mmethot has joined #openstack-barbican14:12
*** moguimar has quit IRC14:37
*** moguimar has joined #openstack-barbican14:37
*** moguimar has quit IRC14:38
*** moguimar has joined #openstack-barbican14:44
*** moguimar has quit IRC14:45
*** moguimar has joined #openstack-barbican14:47
*** velizarx has joined #openstack-barbican14:53
*** velizarx has quit IRC15:09
ade_leedave-mccowan, redrobot we gotta figure something out for barbican gates ..15:09
ade_leedave-mccowan, stuff keeps failing randomly and its not clear how to get a passing set through ..15:09
*** Luzi has quit IRC15:10
ade_leedave-mccowan, redrobot I'm proposing that we make some of the gates non-voting to get some of the backlog cleared15:12
ade_leedave-mccowan, redrobot we'd like to merge to fix the dogtag gate15:12
*** xek has joined #openstack-barbican15:15
*** velizarx has joined #openstack-barbican15:15
dave-mccowanthat's one's already merging.   ade_lee15:17
ade_leedave-mccowan, it is?15:18
ade_leedave-mccowan, its trying to merge -- but keeps getting tripped up15:18
dave-mccowanit's in the queue
dave-mccowanah..  got it, but it will fail again.15:18
dave-mccowanso we need a different patch to make dogtag non-voting.15:19
redrobotIt's the grenade job that keeps failing15:19
redrobotI think the patch fixes the dogtag issue15:19
ade_leedave-mccowan, I'm not worried about the dogtag one -- the patch I have make it work if we can get it merged15:20
ade_leeright - grenade and the devstack functional test fail intermittently15:20
ade_leeie. one passes one time round then fails the next ..15:20
ade_leeare we running into threading issues?15:21
ade_leebrb- coffee ..15:22
redrobotThis is the grenade error: Could not stat /dev/vdb --- No such file or directory15:23
openstackgerritMerged openstack/barbican-specs master: correct wrong word
openstackgerritMerged openstack/barbican-specs master: Replacing the HTTP protocal with HTTPS
*** mmethot has quit IRC15:30
*** mmethot_ has joined #openstack-barbican15:30
ade_leeredrobot, right - maybe a timing thing?15:39
*** abishop has joined #openstack-barbican15:41
*** mmethot_ has quit IRC15:43
*** mmethot_ has joined #openstack-barbican15:43
*** mmethot_ has quit IRC15:46
*** mmethot_ has joined #openstack-barbican15:47
*** mmethot_ has quit IRC15:49
*** mmethot_ has joined #openstack-barbican15:50
ade_leeredrobot, dave-mccowan asked abishop to take a look at the grenade error as it seems to be volume attach related.15:54
ade_leeredrobot, dave-mccowan as for the barbican-devstack-functional error -- any idea here?15:54
redrobotade_lee, have not seen that one15:55
ade_leeredrobot, dave-mccowan we've seen errors like that in the past related to multiple threads running at the same time ..15:56
ade_leethat particular test isn't the issue as much as runnning multiple tests and doing things like trying to count entries15:56
ade_leeor go from the middle one ..15:56
ade_leeare we running serially?15:57
*** livelace has quit IRC16:00
redrobotade_lee, I'm not entirely sure about what command is actually run at the gate16:08
redrobotade_lee, also, it gives me the sads to see our gates are "legacy" :-\16:08
*** mmethot has joined #openstack-barbican16:18
*** pcaruana has quit IRC16:20
*** mmethot has quit IRC16:21
*** mmethot has joined #openstack-barbican16:22
*** mmethot_ has quit IRC16:22
abishopade_lee: something is screwy with the test's timing16:26
abishopcinder reports the volume is in-use at
redrobotRead timed out.16:26
abishopbut then see the logs at
abishop"Retry after 2|3|4 seconds" ?!? timestampts don't corroborate16:27
abishopfinal death at
abishopbut seems to me like enough time has not been given for the hypervisor to propagate the new device, and for the instance to detect its presence16:28
*** mmethot_ has joined #openstack-barbican16:29
ade_leeabishop, yeah - that sounds reasonable16:31
abishopade_lee: totally unfamiliar with these tests, but retry delay seems broken16:31
ade_leeabishop, I would have suspected timing .. so maybe we just need to do a sleep before doing the stat ..16:31
ade_leeabishop, let me pull up where the tests are ..16:32
abishopade_lee: what controls the "Retry after N seconds" ? that seems to not be working16:32
*** _mmethot_ has joined #openstack-barbican16:32
*** mmethot_ has quit IRC16:32
*** mmethot has quit IRC16:33
ade_leeabishop, its the test at the end of the file16:35
ade_leeabishop, we're doing self.attach_detach()16:35
ade_leeline 32616:37
*** mmethot_ has joined #openstack-barbican16:37
*** _mmethot_ has quit IRC16:38
ade_leeabishop, looks like we attach and wait for the volume to be "in-use"16:38
ade_leeabishop, maybe we need to wait for something else?16:38
ade_leeredrobot, right ?  but what is that?16:41
abishopade_lee: I think so; "in-use" means cinder things nova has the volume attached, but it takes time for the hypervisor to add the attachment to the instance, and for the instance to detect its presense16:41
abishopI think a nova person should comment on how all that works16:41
ade_leeabishop, OK - thanks - let me see if I can rustle one up :)16:42
ade_leeredrobot, dave-mccowan suggested we skip the paging tests -- there have not been bugs there for years, and quite a few false positives (esp when things get busy)16:44
*** pcaruana has joined #openstack-barbican17:15
*** dave-mccowan has quit IRC18:26
*** dave-mccowan has joined #openstack-barbican18:32
*** salmankhan has quit IRC18:39
*** velizarx has quit IRC18:41
*** remi_ness has joined #openstack-barbican18:47
*** remi_ness has quit IRC18:54
*** pcaruana has quit IRC19:08
*** remi_ness has joined #openstack-barbican19:25
*** remi_ness has quit IRC20:05
*** whoami-rajat has quit IRC20:40
*** AB2019_ has joined #openstack-barbican21:04
*** AB2019 has quit IRC21:11
*** AB2019_ is now known as AB201921:11
redrobotade_lee, trying to get acquainted with the barbican-tempest-plugin.  I think maybe we should skip that encryption test to unclog the gate?21:15
ade_leeredrobot, or make it non-voting -- there may be other repos that use this21:16
ade_leeredrobot, talking to nova folks but nothing seems obvious21:16
redrobotade_lee, Ok, I'll update your patch to make the grenade test non-voting for now21:18
ade_leeredrobot, yeah and also skip the paging tests21:18
ade_leein the barbican tests21:18
ade_leethose seem to be failing intermittently21:18
redrobotade_lee, ack21:18
ade_leeredrobot, hopefully that clears and we can get through a few of the waiting patches21:20
*** xek has quit IRC21:28
ade_leeredrobot, you uploading the new patch?22:07
openstackgerritAde Lee proposed openstack/barbican master: Work with 389-ds-base-
openstackgerritAde Lee proposed openstack/barbican master: Work with 389-ds-base-
*** mmethot_ has quit IRC23:14
*** raildo has quit IRC23:28
ade_leeredrobot, wtf -- ?23:33
redrobotade_lee, maybe gates are cursed?23:38
ade_leeredrobot, do you have any idea what that error message means?23:45
ade_leeredrobot, btw -- still have that appRole review to do ..23:46
redrobotade_lee, looks like the pep8 tool doesn't like comments without spaces after the #23:49
redrobotade_lee, lines 95 and 96 on functionaltests/api/base.py23:50
redrobotade_lee, I can fix 'em23:50
openstackgerritDouglas Mendizábal proposed openstack/barbican master: Workaround for failing gates

Generated by 2.15.3 by Marius Gedminas - find it at!