Bug #23265
closedFAIL: test_get (tasks.mgr.dashboard_v2.test_cluster_configuration.ClusterConfigurationTest)
0%
Description
The dashboard_v2 API tests became broken in master (sha-1 480f078c147792c6bb234ee0c0b5ab9dbead5b05)
2018-03-07 14:52:49,681.681 INFO:tasks.mgr.dashboard_v2.helper:request POST to http://rdias-suse-laptop.rdias.home.pt:7789//api/auth 2018-03-07 14:52:49,957.957 INFO:tasks.mgr.dashboard_v2.helper:request GET to http://rdias-suse-laptop.rdias.home.pt:7789//api/cluster_conf/admin_socket 2018-03-07 14:52:49,977.977 INFO:__main__:test_get (tasks.mgr.dashboard_v2.test_cluster_configuration.ClusterConfigurationTest) ... FAIL 2018-03-07 14:52:49,977.977 INFO:__main__:run args=['./bin/ceph', 'log', 'Ended test tasks.mgr.dashboard_v2.test_cluster_configuration.ClusterConfigurationTest.test_get'] 2018-03-07 14:52:49,977.977 INFO:__main__:Running ['./bin/ceph', 'log', 'Ended test tasks.mgr.dashboard_v2.test_cluster_configuration.ClusterConfigurationTest.test_get'] 2018-03-07 14:52:51,152.152 INFO:__main__:Stopped test: test_get (tasks.mgr.dashboard_v2.test_cluster_configuration.ClusterConfigurationTest) in 2.902976s 2018-03-07 14:52:51,152.152 INFO:__main__: 2018-03-07 14:52:51,153.153 INFO:__main__:====================================================================== 2018-03-07 14:52:51,153.153 INFO:__main__:FAIL: test_get (tasks.mgr.dashboard_v2.test_cluster_configuration.ClusterConfigurationTest) 2018-03-07 14:52:51,153.153 INFO:__main__:---------------------------------------------------------------------- 2018-03-07 14:52:51,153.153 INFO:__main__:Traceback (most recent call last): 2018-03-07 14:52:51,153.153 INFO:__main__: File "/home/rdias/Work/ceph/qa/tasks/mgr/dashboard_v2/helper.py", line 23, in decorate 2018-03-07 14:52:51,153.153 INFO:__main__: return func(self, *args, **kwargs) 2018-03-07 14:52:51,153.153 INFO:__main__: File "/home/rdias/Work/ceph/qa/tasks/mgr/dashboard_v2/test_cluster_configuration.py", line 43, in test_get 2018-03-07 14:52:51,153.153 INFO:__main__: self.assertStatus(200) 2018-03-07 14:52:51,153.153 INFO:__main__: File "/home/rdias/Work/ceph/qa/tasks/mgr/dashboard_v2/helper.py", line 119, in assertStatus 2018-03-07 14:52:51,153.153 INFO:__main__: self.assertEqual(self._resp.status_code, status) 2018-03-07 14:52:51,153.153 INFO:__main__:AssertionError: 500 != 200 2018-03-07 14:52:51,153.153 INFO:__main__: 2018-03-07 14:52:51,154.154 INFO:__main__:----------------------------------------------------------------------
Updated by Ricardo Dias about 6 years ago
To reproduce this bug:
$ git checkout 480f078c147792c6bb234ee0c0b5ab9dbead5b05 $ cd src/pybind/mgr/dashboard_v2/ $ ./run-backend-api-tests.sh
Updated by Sebastian Wagner about 6 years ago
Here is the PR: https://github.com/ceph/ceph/pull/20782
Updated by Sebastian Wagner about 6 years ago
Some details. Before my commit,
1. md_config_t::config_options
expected Option::dump
to start an object
section. This was no longer the case.
2. Therefore md_config_t::config_options
did return a long array of things.
3. Therefore ActivePyModules::get_python("config_objects") returned garbage
raised an `IndexError` when filtering for a level.
4. Therefore @src/pybind/mgr/dashboard_v2/controllers/cluster_configuration.py:ClusterConfiguration#list
5. Therefore dashboard_v2
returned an Internal Server Error.
6. Therefore The test in /ceph/qa/tasks/mgr/dashboard_v2/test_cluster_configuration.py, line 43, in test_get
failed.
Updated by Sebastian Wagner about 6 years ago
Turned out, this issue consists of two parts: the first one is described above, and the second one was a rename of "developer" to "dev", which was fixed by removing unused code.
Updated by Ricardo Dias about 6 years ago
Sebastian Wagner wrote:
... the second one was a rename of "developer" to "dev", which was fixed by removing unused code.
I don't understand how the issue is fixed just by removing unused code, the "rename" should break the dashboard frontend, right?
Updated by Sebastian Wagner about 6 years ago
I don't understand how the issue is fixed just by removing unused code,
The test failure was fixed by removing unused code. The bug is still present in the frontend code.
Updated by Ernesto Puerta about 3 years ago
- Project changed from mgr to Dashboard
- Category changed from 132 to General