Stephen Finucane b3cb85f112 tests: Improve logging for executed commands
We're seeing failures in a recently added tests,
'ServerTests.test_server_add_remove_volume' from
'openstackclient/tests/functional/compute/v2/test_server.py'. These
failures are likely the result of slow CI nodes, but we don't have
enough information in the CI logs to debug them. Starting logging the
various commands executed in tests so that we can see these logs if and
when tests fail.

Change-Id: I4584dc5e6343fe8c8544431a527d8c3c7e7b3c5b
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
2021-12-09 17:23:32 +00:00

169 lines
5.9 KiB
Python

# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
import logging
import os
import shlex
import subprocess
from tempest.lib.cli import output_parser
from tempest.lib import exceptions
import testtools
ADMIN_CLOUD = os.environ.get('OS_ADMIN_CLOUD', 'devstack-admin')
LOG = logging.getLogger(__name__)
def execute(cmd, fail_ok=False, merge_stderr=False):
"""Executes specified command for the given action."""
LOG.debug('Executing: %s', cmd)
cmdlist = shlex.split(cmd)
stdout = subprocess.PIPE
stderr = subprocess.STDOUT if merge_stderr else subprocess.PIPE
proc = subprocess.Popen(cmdlist, stdout=stdout, stderr=stderr)
result_out, result_err = proc.communicate()
result_out = result_out.decode('utf-8')
LOG.debug('stdout: %s', result_out)
LOG.debug('stderr: %s', result_err)
if not fail_ok and proc.returncode != 0:
raise exceptions.CommandFailed(
proc.returncode, cmd, result_out, result_err,
)
return result_out
class TestCase(testtools.TestCase):
@classmethod
def openstack(cls, cmd, cloud=ADMIN_CLOUD, fail_ok=False):
"""Executes openstackclient command for the given action
NOTE(dtroyer): There is a subtle distinction between pasing
cloud=None and cloud='': for compatibility reasons passing
cloud=None continues to include the option '--os-auth-type none'
in the command while passing cloud='' omits the '--os-auth-type'
option completely to let the default handlers be invoked.
"""
if cloud is None:
# Execute command with no auth
return execute(
'openstack --os-auth-type none ' + cmd,
fail_ok=fail_ok
)
elif cloud == '':
# Execute command with no auth options at all
return execute(
'openstack ' + cmd,
fail_ok=fail_ok
)
else:
# Execure command with an explicit cloud specified
return execute(
'openstack --os-cloud=' + cloud + ' ' + cmd,
fail_ok=fail_ok
)
@classmethod
def is_service_enabled(cls, service, version=None):
"""Ask client cloud if service is available
:param service: The service name or type. This should be either an
exact match to what is in the catalog or a known official value or
alias from service-types-authority
:param version: Optional version. This should be a major version, e.g.
'2.0'
:returns: True if the service is enabled and optionally provides the
specified API version, else False
"""
ret = cls.openstack(
f'versions show --service {service} -f value -c Version'
).splitlines()
if version:
return version in ret
return bool(ret)
@classmethod
def is_extension_enabled(cls, alias):
"""Ask client cloud if extension is enabled"""
return alias in cls.openstack('extension list -f value -c Alias')
@classmethod
def get_openstack_configuration_value(cls, configuration):
opts = cls.get_opts([configuration])
return cls.openstack('configuration show ' + opts)
@classmethod
def get_opts(cls, fields, output_format='value'):
return ' -f {0} {1}'.format(output_format,
' '.join(['-c ' + it for it in fields]))
@classmethod
def assertOutput(cls, expected, actual):
if expected != actual:
raise Exception(expected + ' != ' + actual)
@classmethod
def assertInOutput(cls, expected, actual):
if expected not in actual:
raise Exception(expected + ' not in ' + actual)
@classmethod
def assertsOutputNotNone(cls, observed):
if observed is None:
raise Exception('No output observed')
def assert_table_structure(self, items, field_names):
"""Verify that all items have keys listed in field_names."""
for item in items:
for field in field_names:
self.assertIn(field, item)
def assert_show_fields(self, show_output, field_names):
"""Verify that all items have keys listed in field_names."""
# field_names = ['name', 'description']
# show_output = [{'name': 'fc2b98d8faed4126b9e371eda045ade2'},
# {'description': 'description-821397086'}]
# this next line creates a flattened list of all 'keys' (like 'name',
# and 'description' out of the output
all_headers = [item for sublist in show_output for item in sublist]
for field_name in field_names:
self.assertIn(field_name, all_headers)
def parse_show_as_object(self, raw_output):
"""Return a dict with values parsed from cli output."""
items = self.parse_show(raw_output)
o = {}
for item in items:
o.update(item)
return o
def parse_show(self, raw_output):
"""Return list of dicts with item values parsed from cli output."""
items = []
table_ = output_parser.table(raw_output)
for row in table_['values']:
item = {}
item[row[0]] = row[1]
items.append(item)
return items
def parse_listing(self, raw_output):
"""Return list of dicts with basic item parsed from cli output."""
return output_parser.listing(raw_output)