Add specific common tripleo container checks , and nova services,
and fix podman client bug Change-Id: Ie05810c93e25077ff9e4b8cc9ab128288af710c7
This commit is contained in:
pinikomarov
parent
5fa2b48619
commit
a1b21aec06
|
|
@ -113,12 +113,16 @@ class PodmanClientFixture(tobiko.SharedFixture):
|
|||
|
||||
try:
|
||||
podman_remote_socket = self.discover_podman_socket()
|
||||
podman_remote_socket_uri = 'unix:/tmp/podman.sock'
|
||||
username = self.ssh_client.connect_parameters['username']
|
||||
host = self.ssh_client.connect_parameters["hostname"]
|
||||
socket = podman_remote_socket
|
||||
podman_remote_socket_uri = \
|
||||
'unix:/tmp/podman.sock_{}'.format(host)
|
||||
|
||||
remote_uri = 'ssh://{username}@{host}{socket}'.format(
|
||||
username=self.ssh_client.connect_parameters['username'],
|
||||
host=self.ssh_client.connect_parameters["hostname"],
|
||||
socket=podman_remote_socket)
|
||||
username=username,
|
||||
host=host,
|
||||
socket=socket)
|
||||
|
||||
client = podman.Client(uri=podman_remote_socket_uri,
|
||||
remote_uri=remote_uri,
|
||||
|
|
|
|||
|
|
@ -10,6 +10,7 @@ from tobiko.tripleo import pacemaker
|
|||
from tobiko.tripleo import processes
|
||||
from tobiko.tripleo import containers
|
||||
from tobiko.tripleo import neutron
|
||||
from tobiko.tripleo import nova
|
||||
from tobiko.openstack import stacks
|
||||
import tobiko
|
||||
|
||||
|
|
@ -18,14 +19,14 @@ def nodes_health_check():
|
|||
# this method will be changed in future commit
|
||||
check_pacemaker_resources_health()
|
||||
check_overcloud_processes_health()
|
||||
nova.check_nova_services_health()
|
||||
neutron.check_neutron_agents_health()
|
||||
containers.assert_all_tripleo_containers_running()
|
||||
# create a uniq stack
|
||||
check_vm_create(stack_name='stack{}'.format(random.randint(0, 10000)))
|
||||
|
||||
# TODO:
|
||||
# Test existing created serverstest_controller_containers
|
||||
# ServerStackResourcesTest().test_server_create()
|
||||
# Add specific container checks
|
||||
# Test existing created servers
|
||||
|
||||
|
||||
# check vm create with ssh and ping checks
|
||||
|
|
|
|||
|
|
@ -15,7 +15,7 @@ LOG = log.getLogger(__name__)
|
|||
def container_runtime():
|
||||
"""check what container runtime is running
|
||||
and return a handle to it"""
|
||||
|
||||
# TODO THIS LOCKS SSH CLIENT TO CONTROLLER
|
||||
ssh_client = topology.list_openstack_nodes(group='controller')[
|
||||
0].ssh_client
|
||||
if docker.is_docker_running(ssh_client=ssh_client):
|
||||
|
|
@ -27,7 +27,7 @@ def container_runtime():
|
|||
container_runtime_type = container_runtime()
|
||||
|
||||
|
||||
def list_node_containers(client=None):
|
||||
def list_node_containers(client):
|
||||
"""returns a list of containers and their run state"""
|
||||
|
||||
if container_runtime_type == podman:
|
||||
|
|
@ -70,6 +70,94 @@ def list_containers(group=None):
|
|||
return containers_list
|
||||
|
||||
|
||||
def assert_containers_running(group, excpected_containers):
|
||||
|
||||
"""assert that all containers specified in the list are running
|
||||
on the specified openstack group(controller or compute etc..)"""
|
||||
|
||||
failures = []
|
||||
|
||||
openstack_nodes = topology.list_openstack_nodes(group=group)
|
||||
for node in openstack_nodes:
|
||||
container_client = get_container_client(node.ssh_client)
|
||||
node_containers = list_node_containers(client=container_client)
|
||||
containers_list_df = pandas.DataFrame(
|
||||
get_container_states_list(node_containers),
|
||||
columns=['container_host', 'container_name', 'container_state'])
|
||||
# check that the containers are present
|
||||
LOG.info('node: {} containers list : {}'.format(
|
||||
node.name, containers_list_df.to_string(index=False)))
|
||||
for container in excpected_containers:
|
||||
# get container attrs dataframe
|
||||
container_attrs = containers_list_df.query(
|
||||
'container_name == "{}"'.format(container))
|
||||
# check if the container exists
|
||||
LOG.info('checking container: {}'.format(container))
|
||||
if container_attrs.empty:
|
||||
failures.append(
|
||||
'expected container {} not found on node {} ! : \n\n'.
|
||||
format(container, node.name))
|
||||
# if container exists, check it is running
|
||||
else:
|
||||
container_state = \
|
||||
container_attrs.container_state.values.item()
|
||||
if not container_state == 'running':
|
||||
failures.append(
|
||||
'expected container {} is not running on node {} , '
|
||||
'its state is {}! : \n\n'.format(container,
|
||||
node.name,
|
||||
container_state))
|
||||
|
||||
if failures:
|
||||
tobiko.fail('container states mismatched:\n{!s}', '\n'.join(failures))
|
||||
else:
|
||||
LOG.info('All tripleo common containers are in running state! ')
|
||||
|
||||
|
||||
def assert_all_tripleo_containers_running():
|
||||
"""check that all common tripleo containers are running
|
||||
param: group controller or compute , check containers
|
||||
sets in computes or controllers"""
|
||||
|
||||
common_controller_tripleo_containers = ['cinder_api', 'cinder_api_cron',
|
||||
'cinder_scheduler', 'clustercheck',
|
||||
'glance_api', 'heat_api',
|
||||
'heat_api_cfn',
|
||||
'heat_api_cron', 'heat_engine',
|
||||
'horizon', 'iscsid', 'keystone',
|
||||
'logrotate_crond', 'memcached',
|
||||
'neutron_api', 'nova_api',
|
||||
'nova_api_cron', 'nova_conductor',
|
||||
'nova_metadata', 'nova_scheduler',
|
||||
'nova_vnc_proxy',
|
||||
'swift_account_auditor',
|
||||
'swift_account_reaper',
|
||||
'swift_account_replicator',
|
||||
'swift_account_server',
|
||||
'swift_container_auditor',
|
||||
'swift_container_replicator',
|
||||
'swift_container_server',
|
||||
'swift_container_updater',
|
||||
'swift_object_auditor',
|
||||
'swift_object_expirer',
|
||||
'swift_object_replicator',
|
||||
'swift_object_server',
|
||||
'swift_object_updater',
|
||||
'swift_proxy', 'swift_rsync']
|
||||
|
||||
common_compute_tripleo_containers = ['iscsid', 'logrotate_crond',
|
||||
'nova_compute', 'nova_libvirt',
|
||||
'nova_migration_target',
|
||||
'nova_virtlogd']
|
||||
|
||||
for group, group_containers in [('controller',
|
||||
common_controller_tripleo_containers),
|
||||
('compute',
|
||||
common_compute_tripleo_containers)]:
|
||||
assert_containers_running(group, group_containers)
|
||||
# TODO: need to address OSP-version specific containers here.
|
||||
|
||||
|
||||
def comparable_container_keys(container):
|
||||
"""returns the tuple : 'container_host','container_name',
|
||||
'container_state'
|
||||
|
|
|
|||
|
|
@ -11,7 +11,7 @@ def check_neutron_agents_health():
|
|||
|
||||
for agent in agents['agents']:
|
||||
if not agent['alive']:
|
||||
failures.append('failed agent: {}'.format(agent))
|
||||
failures.append('failed agent: {}\n\n'.format(agent))
|
||||
|
||||
if failures:
|
||||
tobiko.fail(
|
||||
|
|
|
|||
|
|
@ -0,0 +1,18 @@
|
|||
from __future__ import absolute_import
|
||||
|
||||
import tobiko
|
||||
from tobiko.openstack import nova
|
||||
|
||||
|
||||
def check_nova_services_health():
|
||||
failures = []
|
||||
nova_client = nova.get_nova_client()
|
||||
services = nova_client.services.list()
|
||||
|
||||
for service in services:
|
||||
if not service.state == 'up':
|
||||
failures.append('failed service: {}\n\n'.format(vars(service)))
|
||||
|
||||
if failures:
|
||||
tobiko.fail(
|
||||
'nova agents are unhealthy:\n{!s}', '\n'.join(failures))
|
||||
Loading…
Reference in New Issue