Earlier  
Later
New Search
Posted Nick Remark
#openstack-sdks - 2017-07-12
07:53:09 amotoki when both os-client-config and osc-lib are downgraded, osc unit test succeeded.
07:53:22 amotoki is anyone investigating the detail?
07:53:47 amotoki Good: osc-lib===1.6.0 and os-client-config===1.27.0
07:53:55 amotoki Bad: osc-lib===1.7.0 and os-client-config===1.28.0
08:06:16 openstackgerrit Rui Chen proposed openstack/python-openstackclient master: Clean up the changes of os.environ in functional tests https://review.openstack.org/471341
08:06:22 openstackgerrit Rui Chen proposed openstack/python-openstackclient master: Rework server migrate to support cold and live https://review.openstack.org/460059
08:06:31 openstackgerrit Rui Chen proposed openstack/python-openstackclient master: [Compute]Make column content readable for both human and machine https://review.openstack.org/445794
08:17:04 amotoki the failure patterns related to osc-lib 1.7.0, osc-lib 1.7.0 is good. osc volume find_resource unit tests look bad.
08:18:10 amotoki RuiChen: are you looking into the ut failure?
08:23:53 RuiChen amotoki: not yet, that impact all the volume unit tests?
08:24:25 amotoki RuiChen: it only affects openstackclient.tests.unit.volume.test_find_resource
08:25:16 RuiChen amotoki: let me try it in my local env with libs
08:25:17 amotoki RuiChen: looking at osc-lib change, it drops unnecessary try-except. according to osc-lib comments, it was a dirty-hack to hide test errors
08:25:51 amotoki RuiChen: i am now looking at os-client-config failure mode to determine we need to block it in g-r.
08:50:35 amotoki I filed bugs on the ut failures: bug 1703782 and bug 1703783 for tracking them.
08:50:36 openstack bug 1703782 in python-openstackclient "OSC unit test volume.test_find_resource fails with osc-lib 1.7.0" [Undecided,New] https://launchpad.net/bugs/1703782
08:50:37 openstack bug 1703783 in python-openstackclient "OSC unit fails with os-client-config 1.27.0" [Undecided,New] https://launchpad.net/bugs/1703783
09:16:17 amotoki hm.... openstack command calls OpenStackConfig.__init__() TWICE... why?
09:17:11 amotoki it hits os-client-config check and causes the UT failures. tricky :(
09:18:32 amotoki I have no time to investigate more on OSC UT failure today.
09:32:51 amotoki RuiChen: i identified the cause of the failure related to os-client-config. will upload a patch soon.
09:50:43 openstackgerrit Akihiro Motoki proposed openstack/python-openstackclient master: Fix os-client-config related unit tests for occ 1.28.0 https://review.openstack.org/482880
10:01:46 openstackgerrit Akihiro Motoki proposed openstack/python-openstackclient master: Fix os-client-config related unit tests for occ 1.28.0 https://review.openstack.org/482880
10:06:31 openstackgerrit Akihiro Motoki proposed openstack/python-openstackclient master: Fix os-client-config related unit tests for occ 1.28.0 https://review.openstack.org/482880
10:08:07 RuiChen amotoki: it's great, thank you pushing the patch, about the volume ut failures, I think it's related with https://review.openstack.org/#/c/476523/ changes
10:08:52 amotoki RuiChen: it is not related to volume ut failures. it's related to the new os-client-config.
10:09:17 amotoki RuiChen: ah. I misunderstood your comment.
10:09:40 amotoki RuiChen: all are delimited by a comma. I got confused
10:10:12 RuiChen I think your patch to fix the openstackclient.tests.unit.integ.cli.test_shell.TestIntegShellCliPrecedenceOCC failures, right?
10:10:53 amotoki RuiChen: right
10:11:11 amotoki also skips volume find_resource to pass py27.
10:13:07 RuiChen amotoki: ah, I got it, thank you
10:18:50 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add min_version and max_version to adapter constructors https://review.openstack.org/482742
10:18:51 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Expose interface as a ListOpt for config files https://review.openstack.org/482743
10:18:52 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add paragraph clarifying major and micro versions https://review.openstack.org/482710
10:18:53 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add ability to request a microversion in a request https://review.openstack.org/482746
10:18:54 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Update docs and add a release note https://review.openstack.org/477566
10:18:55 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Remove kwargs from adapter.get_endpoint_data https://review.openstack.org/482744
10:18:56 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add support for next_min_version and not_before https://review.openstack.org/482745
11:35:21 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Fix typo from earlier docs patch https://review.openstack.org/482925
12:13:48 openstackgerrit Akihiro Motoki proposed openstack/python-openstackclient master: Fix unit test failures related to new os-client-config and osc-lib https://review.openstack.org/482880
12:14:35 amotoki RuiChen: thanks for your finding in UT failure around osc-lib. I imported your suggestion in the new patch set above.
12:26:33 openstackgerrit Akihiro Motoki proposed openstack/python-openstackclient master: Fix unit test failures related to new os-client-config and osc-lib https://review.openstack.org/482880
12:47:45 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add min_version and max_version to adapter constructors https://review.openstack.org/482742
12:47:46 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Expose interface as a ListOpt for config files https://review.openstack.org/482743
12:47:47 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add paragraph clarifying major and micro versions https://review.openstack.org/482710
12:47:48 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add ability to request a microversion in a request https://review.openstack.org/482746
12:47:49 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Update docs and add a release note https://review.openstack.org/477566
12:47:50 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Remove kwargs from adapter.get_endpoint_data https://review.openstack.org/482744
12:47:51 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add support for next_min_version and not_before https://review.openstack.org/482745
13:16:31 GreenBlood I'm looking at the v2/v3 API for identity managment and while in v2 I see that querying https://[endpoint]/v2 replies with version details, I don't see a corresponding function on v3. When I query my provider with /v3, the reply I get is "Api does not exist". Is this normal ?
13:16:46 GreenBlood ( https://developer.openstack.org/api-ref/identity/v3/ and https://developer.openstack.org/api-ref/identity/v2/?expanded=show-version-details-detail )
13:17:35 efried GreenBlood Doubt that's okay. mordred ^^
13:20:46 GreenBlood I fired up a devstack and launched a curl on the v3 url and the reply is pretty much the same as v2
13:20:56 GreenBlood so i guess this answers my question
13:21:03 GreenBlood thanks anyways :D
15:23:05 openstackgerrit Monty Taylor proposed openstack/api-wg master: Add document describing consuming version discovery https://review.openstack.org/459405
20:44:43 dhellmann sbezverk : how long do those kolla-kubernetes tests take? do I need anything special installed to run them?
20:45:27 sbezverk dhellmann it should not take longer than 1 hour
20:45:35 dhellmann "tox -e py27" returns immediately without running any tests, so *how* do I run them?
20:46:03 dhellmann ok, I can't tie up my box for that long, so I'll submit a patch on top of yours to get some more data
20:46:30 sbezverk dhellmann tox tests will take minutes
20:46:46 sbezverk the whole job takes about 40 minutes
20:46:48 dhellmann oh, wait, I was in the wrong window. duh.
20:47:53 dhellmann ouch, what's it installing into my home directory?
20:48:23 sbezverk some dependencies
20:48:35 sbezverk we usually start it in virtenv
20:48:43 dhellmann well, I ran "tox"
20:48:50 dhellmann it created ~/.helm ?
20:48:54 dhellmann I'm going to set up a vm I can throw away
20:49:45 sbezverk dhellmann : ok helm charts are what actually tested
20:52:53 dhellmann ok, running
20:52:57 dhellmann and I reproduced the error
20:53:35 sbezverk yeah it is 100% consistent :( so our gates are not functional
20:54:20 dhellmann sbezverk : can I pass a test name in to have only that test run?
20:55:13 dhellmann I see some errors in the part of the output where it is installing or running helm, are you aware of those?
20:55:17 dhellmann they are not likely to be related
20:56:18 sbezverk dhellmann : yes, that helm error can be ignored
20:56:24 dhellmann ok
20:57:00 dhellmann something is setting the command object's app value to a string instead of an application object
20:57:04 dhellmann let me see if I can figure out where that is
20:57:42 dhellmann oh, it's right there in the constructor
20:58:38 dhellmann sbezverk : this line is wrong: http://git.openstack.org/cgit/openstack/kolla-kubernetes/tree/kolla_kubernetes/tests/test_templates.py#n70
20:58:51 dhellmann the first argument to the Command object needs to be a subclass of cliff.app.App
20:59:28 dhellmann the new use of hooks is triggering the need for that in a path that didn't exist before, but it has always been the expectation
21:00:54 sbezverk ok, I will reach out to the developer who worked on this
21:04:58 dhellmann sbezverk : https://review.openstack.org/483097 fixes it
21:05:04 sbezverk thank you very much for looking into it
21:05:50 dhellmann sbezverk : I'm glad I could help
21:19:03 openstackgerrit Merged openstack/python-openstackclient master: Fix unit test failures related to new os-client-config and osc-lib https://review.openstack.org/482880
21:19:37 sbezverk dhellmann : https://review.openstack.org/#/c/483099/1 is not good as all our charts are stored there
21:20:05 sbezverk and they are part of repo
21:20:30 sbezverk it probably created in .helm in your home directory
21:20:43 sbezverk but that is outside of repo
21:29:13 dhellmann sbezverk : ok. I just posted a new one that ignores the lock files that I found in that directory after running the tests
21:29:57 dhellmann sbezverk : if that's no good, I'll stop until I know what helm even is :-)
21:34:28 sbezverk dhellmann : lock files clean up is good, never had time to address it, so thank you for doing it !!
21:36:21 dhellmann sbezverk : no problem :-)
22:50:09 openstackgerrit Steve Martinelli proposed openstack/python-openstackclient master: Add 'data_plane_status' option to Port classes https://review.openstack.org/458626
22:50:17 openstackgerrit Steve Martinelli proposed openstack/python-openstackclient master: image-list should support filters 'name','status' https://review.openstack.org/479670

Earlier   Later