openstack
/
ironic
Code Issues Proposed changes

Merge "Fix idrac-wsman RAID step async error handling"

This commit is contained in:
Zuul 2020-11-12 19:08:59 +00:00 committed by Gerrit Code Review
parent 887f263c26 23951f4b44
commit b5932bc6bf
3 changed files with 44 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
ironic/drivers/modules/drac/raid.py
View File

@ -1491,13 +1491,18 @@ class DracWSManRAID(base.RAIDInterface):
node.save()
def _set_failed(self, task, config_job):
LOG.error("RAID configuration job failed for node %(node)s. "
"Failed config job: %(config_job_id)s. "
"Message: '%(message)s'.",
{'node': task.node.uuid, 'config_job_id': config_job.id,
'message': config_job.message})
task.node.last_error = config_job.message
task.process_event('fail')
error_msg = (_("Failed config job: %(config_job_id)s. "
"Message: '%(message)s'.") %
{'config_job_id': config_job.id,
'message': config_job.message})
log_msg = (_("RAID configuration job failed for node %(node)s. "
"%(error)s") %
{'node': task.node.uuid, 'error': error_msg})
if task.node.clean_step:
LOG.error(log_msg)
manager_utils.cleaning_error_handler(task, error_msg)
else:
manager_utils.deploying_error_handler(task, log_msg, error_msg)
def _resume(self, task):
raid_common.update_raid_info(

30
ironic/tests/unit/drivers/modules/drac/test_periodic_task.py
View File

@ -203,9 +203,11 @@ class DracPeriodicTaskTestCase(db_base.DbTestCase):
self._test__check_node_raid_jobs_with_completed_job(
mock_notify_conductor_resume)
@mock.patch.object(manager_utils, 'cleaning_error_handler', autospec=True)
@mock.patch.object(drac_common, 'get_drac_client', spec_set=True,
autospec=True)
def test__check_node_raid_jobs_with_failed_job(self, mock_get_drac_client):
def test__check_node_raid_jobs_with_failed_job(
self, mock_get_drac_client, mock_cleaning_error_handler):
# mock node.driver_internal_info and node.clean_step
driver_internal_info = {'raid_config_job_ids': ['42']}
self.node.driver_internal_info = driver_internal_info
@ -232,15 +234,18 @@ class DracPeriodicTaskTestCase(db_base.DbTestCase):
self.assertEqual([],
self.node.driver_internal_info['raid_config_job_ids'])
self.assertEqual({}, self.node.raid_config)
task.process_event.assert_called_once_with('fail')
mock_cleaning_error_handler.assert_called_once_with(task, mock.ANY)
@mock.patch.object(manager_utils, 'deploying_error_handler', autospec=True)
@mock.patch.object(manager_utils, 'cleaning_error_handler', autospec=True)
@mock.patch.object(drac_common, 'get_drac_client', spec_set=True,
autospec=True)
@mock.patch.object(drac_raid.DracRAID, 'get_logical_disks',
spec_set=True, autospec=True)
def _test__check_node_raid_jobs_with_completed_job_already_failed(
self, mock_notify_conductor_resume,
mock_get_logical_disks, mock_get_drac_client):
mock_get_logical_disks, mock_get_drac_client,
mock_cleaning_error_handler, mock_deploying_error_handler):
expected_logical_disk = {'size_gb': 558,
'raid_level': '1',
'name': 'disk 0'}
@ -271,7 +276,12 @@ class DracPeriodicTaskTestCase(db_base.DbTestCase):
self.assertNotIn('raid_config_job_failure',
self.node.driver_internal_info)
self.assertNotIn('logical_disks', self.node.raid_config)
task.process_event.assert_called_once_with('fail')
if self.node.clean_step:
mock_cleaning_error_handler.assert_called_once_with(task, mock.ANY)
else:
mock_deploying_error_handler.assert_called_once_with(task,
mock.ANY,
mock.ANY)
self.assertFalse(mock_notify_conductor_resume.called)
@mock.patch.object(manager_utils, 'notify_conductor_resume_clean',
@ -346,13 +356,16 @@ class DracPeriodicTaskTestCase(db_base.DbTestCase):
self._test__check_node_raid_jobs_with_multiple_jobs_completed(
mock_notify_conductor_resume)
@mock.patch.object(manager_utils, 'deploying_error_handler', autospec=True)
@mock.patch.object(manager_utils, 'cleaning_error_handler', autospec=True)
@mock.patch.object(drac_common, 'get_drac_client', spec_set=True,
autospec=True)
@mock.patch.object(drac_raid.DracRAID, 'get_logical_disks',
spec_set=True, autospec=True)
def _test__check_node_raid_jobs_with_multiple_jobs_failed(
self, mock_notify_conductor_resume,
mock_get_logical_disks, mock_get_drac_client):
mock_get_logical_disks, mock_get_drac_client,
mock_cleaning_error_handler, mock_deploying_error_handler):
expected_logical_disk = {'size_gb': 558,
'raid_level': '1',
'name': 'disk 0'}
@ -387,7 +400,12 @@ class DracPeriodicTaskTestCase(db_base.DbTestCase):
self.assertNotIn('raid_config_job_failure',
self.node.driver_internal_info)
self.assertNotIn('logical_disks', self.node.raid_config)
task.process_event.assert_called_once_with('fail')
if self.node.clean_step:
mock_cleaning_error_handler.assert_called_once_with(task, mock.ANY)
else:
mock_deploying_error_handler.assert_called_once_with(task,
mock.ANY,
mock.ANY)
self.assertFalse(mock_notify_conductor_resume.called)
@mock.patch.object(manager_utils, 'notify_conductor_resume_clean',

8
releasenotes/notes/fix-wsman-raid-async-step-error-handling-f44e2001ac018d12.yaml Normal file
View File

@ -0,0 +1,8 @@
---
fixes:
- |
Fixes ``idrac-wsman`` RAID ``create_configuration`` clean step,
``apply_configuration`` deploy step and ``delete_configuration`` clean and
deploy step to fail correctly in case of error when checking completed
jobs. Before the fix when RAID job failed, then node cleaning or deploying
failed with timeout instead of actual error in clean or deploy step.