Merge pull request #418 from openstack-charmers/issue/417
Log resource cleanup failure and carry on, retry lb resource creation and various other retry fixes
This commit is contained in:
@@ -15,6 +15,7 @@
|
||||
"""Code for configuring octavia-diskimage-retrofit."""
|
||||
|
||||
import logging
|
||||
import tenacity
|
||||
|
||||
import zaza.model
|
||||
|
||||
@@ -39,12 +40,20 @@ def retrofit_amphora_image(unit='octavia-diskimage-retrofit/0',
|
||||
if image_id:
|
||||
params.update({'source-image': image_id})
|
||||
|
||||
# NOTE(fnordahl) ``zaza.model.run_action_on_leader`` fails here,
|
||||
# apparently has to do with handling of subordinates in ``libjuju`` or
|
||||
# ``juju`` itself.
|
||||
action = zaza.model.run_action(
|
||||
unit,
|
||||
'retrofit-image',
|
||||
action_params=params,
|
||||
raise_on_failure=True)
|
||||
# NOTE(fnordahl) the retrofit process involves downloading packages from
|
||||
# the internet and is as such susceptible to random failures due to
|
||||
# internet gremlins.
|
||||
for attempt in tenacity.Retrying(
|
||||
stop=tenacity.stop_after_attempt(3),
|
||||
wait=tenacity.wait_exponential(
|
||||
multiplier=1, min=2, max=10)):
|
||||
with attempt:
|
||||
# NOTE(fnordahl) ``zaza.model.run_action_on_leader`` fails here,
|
||||
# apparently has to do with handling of subordinates in ``libjuju``
|
||||
# or ``juju`` itself.
|
||||
action = zaza.model.run_action(
|
||||
unit,
|
||||
'retrofit-image',
|
||||
action_params=params,
|
||||
raise_on_failure=True)
|
||||
return action
|
||||
|
||||
@@ -18,6 +18,8 @@ import logging
|
||||
import subprocess
|
||||
import tenacity
|
||||
|
||||
from keystoneauth1 import exceptions as keystone_exceptions
|
||||
import octaviaclient.api.v2.octavia
|
||||
import osc_lib.exceptions
|
||||
|
||||
import zaza.openstack.charm_tests.test_utils as test_utils
|
||||
@@ -80,18 +82,62 @@ class LBAASv2Test(test_utils.OpenStackBaseTest):
|
||||
# List of floating IPs created by this test
|
||||
cls.fips = []
|
||||
|
||||
def resource_cleanup(self):
|
||||
"""Remove resources created during test execution."""
|
||||
def _remove_amphorae_instances(self):
|
||||
"""Remove amphorae instances forcefully.
|
||||
|
||||
In some situations Octavia is unable to remove load balancer resources.
|
||||
This helper can be used to remove the underlying instances.
|
||||
"""
|
||||
result = self.octavia_client.amphora_list()
|
||||
for amphora in result.get('amphorae', []):
|
||||
for server in self.nova_client.servers.list():
|
||||
if 'compute_id' in amphora and server.id == amphora[
|
||||
'compute_id']:
|
||||
try:
|
||||
openstack_utils.delete_resource(
|
||||
self.nova_client.servers,
|
||||
server.id,
|
||||
msg="server")
|
||||
except AssertionError as e:
|
||||
logging.warning(
|
||||
'Gave up waiting for resource cleanup: "{}"'
|
||||
.format(str(e)))
|
||||
|
||||
@tenacity.retry(stop=tenacity.stop_after_attempt(3),
|
||||
wait=tenacity.wait_exponential(
|
||||
multiplier=1, min=2, max=10))
|
||||
def resource_cleanup(self, only_local=False):
|
||||
"""Remove resources created during test execution.
|
||||
|
||||
:param only_local: When set to true do not call parent method
|
||||
:type only_local: bool
|
||||
"""
|
||||
for lb in self.loadbalancers:
|
||||
self.octavia_client.load_balancer_delete(lb['id'], cascade=True)
|
||||
try:
|
||||
self.wait_for_lb_resource(
|
||||
self.octavia_client.load_balancer_show, lb['id'],
|
||||
provisioning_status='DELETED')
|
||||
except osc_lib.exceptions.NotFound:
|
||||
pass
|
||||
self.octavia_client.load_balancer_delete(
|
||||
lb['id'], cascade=True)
|
||||
except octaviaclient.api.v2.octavia.OctaviaClientException as e:
|
||||
logging.info('Octavia is unable to delete load balancer: "{}"'
|
||||
.format(e))
|
||||
logging.info('Attempting to forcefully remove amphorae')
|
||||
self._remove_amphorae_instances()
|
||||
else:
|
||||
try:
|
||||
self.wait_for_lb_resource(
|
||||
self.octavia_client.load_balancer_show, lb['id'],
|
||||
provisioning_status='DELETED')
|
||||
except osc_lib.exceptions.NotFound:
|
||||
pass
|
||||
# allow resource cleanup to be run multiple times
|
||||
self.loadbalancers = []
|
||||
for fip in self.fips:
|
||||
self.neutron_client.delete_floatingip(fip)
|
||||
# allow resource cleanup to be run multiple times
|
||||
self.fips = []
|
||||
|
||||
if only_local:
|
||||
return
|
||||
|
||||
# we run the parent resource_cleanup last as it will remove instances
|
||||
# referenced as members in the above cleaned up load balancers
|
||||
super(LBAASv2Test, self).resource_cleanup()
|
||||
@@ -157,6 +203,7 @@ class LBAASv2Test(test_utils.OpenStackBaseTest):
|
||||
'provider': provider,
|
||||
}})
|
||||
lb = result['loadbalancer']
|
||||
self.loadbalancers.append(lb)
|
||||
lb_id = lb['id']
|
||||
|
||||
logging.info('Awaiting loadbalancer to reach provisioning_status '
|
||||
@@ -283,10 +330,25 @@ class LBAASv2Test(test_utils.OpenStackBaseTest):
|
||||
for provider in self.get_lb_providers(self.octavia_client).keys():
|
||||
logging.info('Creating loadbalancer with provider {}'
|
||||
.format(provider))
|
||||
lb = self._create_lb_resources(self.octavia_client, provider,
|
||||
vip_subnet_id, subnet_id,
|
||||
payload_ips)
|
||||
self.loadbalancers.append(lb)
|
||||
final_exc = None
|
||||
# NOTE: we cannot use tenacity here as the method we call into
|
||||
# already uses it to wait for operations to complete.
|
||||
for retry in range(0, 3):
|
||||
try:
|
||||
lb = self._create_lb_resources(self.octavia_client,
|
||||
provider,
|
||||
vip_subnet_id,
|
||||
subnet_id,
|
||||
payload_ips)
|
||||
break
|
||||
except (AssertionError,
|
||||
keystone_exceptions.connection.ConnectFailure) as e:
|
||||
logging.info('Retrying load balancer creation, last '
|
||||
'failure: "{}"'.format(str(e)))
|
||||
self.resource_cleanup(only_local=True)
|
||||
final_exc = e
|
||||
else:
|
||||
raise final_exc
|
||||
|
||||
lb_fp = openstack_utils.create_floating_ip(
|
||||
self.neutron_client, 'ext_net', port={'id': lb['vip_port_id']})
|
||||
|
||||
@@ -495,6 +495,11 @@ class OpenStackBaseTest(BaseCharmTest):
|
||||
self.nova_client.servers,
|
||||
server.id,
|
||||
msg="server")
|
||||
except AssertionError as e:
|
||||
# Resource failed to be removed within the expected time frame,
|
||||
# log this fact and carry on.
|
||||
logging.warning('Gave up waiting for resource cleanup: "{}"'
|
||||
.format(str(e)))
|
||||
except AttributeError:
|
||||
# Test did not define self.RESOURCE_PREFIX, ignore.
|
||||
pass
|
||||
|
||||
@@ -690,14 +690,20 @@ def add_interface_to_netplan(server_name, mac_address):
|
||||
"{}\nserver_name: {}".format(body_value, unit_name,
|
||||
interface, mac_address,
|
||||
server_name))
|
||||
with tempfile.NamedTemporaryFile(mode="w") as netplan_file:
|
||||
netplan_file.write(body_value)
|
||||
netplan_file.flush()
|
||||
model.scp_to_unit(unit_name, netplan_file.name,
|
||||
'/home/ubuntu/60-dataport.yaml', user="ubuntu")
|
||||
run_cmd_mv = "sudo mv /home/ubuntu/60-dataport.yaml /etc/netplan/"
|
||||
model.run_on_unit(unit_name, run_cmd_mv)
|
||||
model.run_on_unit(unit_name, "sudo netplan apply")
|
||||
for attempt in tenacity.Retrying(
|
||||
stop=tenacity.stop_after_attempt(3),
|
||||
wait=tenacity.wait_exponential(
|
||||
multiplier=1, min=2, max=10)):
|
||||
with attempt:
|
||||
with tempfile.NamedTemporaryFile(mode="w") as netplan_file:
|
||||
netplan_file.write(body_value)
|
||||
netplan_file.flush()
|
||||
model.scp_to_unit(
|
||||
unit_name, netplan_file.name,
|
||||
'/home/ubuntu/60-dataport.yaml', user="ubuntu")
|
||||
run_cmd_mv = "sudo mv /home/ubuntu/60-dataport.yaml /etc/netplan/"
|
||||
model.run_on_unit(unit_name, run_cmd_mv)
|
||||
model.run_on_unit(unit_name, "sudo netplan apply")
|
||||
|
||||
|
||||
def configure_gateway_ext_port(novaclient, neutronclient, net_id=None,
|
||||
|
||||
Reference in New Issue
Block a user