Do not call update_device_list in large sets

Ovs-agent can process the ports in large sets, then all
of these ports will have to update DB status or attributes.
But neutron server is centralized. It may have to do
something else, or the database processing can be also
time-consuming. Because of these, it sometimes returns
the RPC timeout exception to ovs-agent. And a fullsync
will be triggered in next rpc loop. The restart time is
becoming longer and longer.

Adds a default step to update the port to reduce
the probability of RPC timeout.

Related-Bug: #1813703
Related-Bug: #1813704
Related-Bug: #1813706
Related-Bug: #1813707

Conflicts:
	neutron/tests/unit/plugins/ml2/test_rpc.py

Change-Id: Ie37f4a4869969e235ce16b73cdfcbdc98626823e
(cherry picked from commit 8408af4f17)
(cherry picked from commit d7d30ea950)
(cherry picked from commit 5d705468de)
changes/91/649691/4
LIU Yulong 4 years ago committed by Vlad Gusev
parent 7ba0d8fb49
commit fa16540d2d

@ -146,10 +146,31 @@ class PluginApi(object):
def update_device_list(self, context, devices_up, devices_down,
agent_id, host, agent_restarted=False):
cctxt = self.client.prepare(version='1.5')
return cctxt.call(context, 'update_device_list',
devices_up=devices_up, devices_down=devices_down,
agent_id=agent_id, host=host,
agent_restarted=agent_restarted)
ret_devices_up = []
failed_devices_up = []
ret_devices_down = []
failed_devices_down = []
step = n_const.RPC_RES_PROCESSING_STEP
devices_up = list(devices_up)
devices_down = list(devices_down)
for i in range(0, max(len(devices_up), len(devices_down)), step):
# Divide-and-conquer RPC timeout
ret = cctxt.call(context, 'update_device_list',
devices_up=devices_up[i:i + step],
devices_down=devices_down[i:i + step],
agent_id=agent_id, host=host,
agent_restarted=agent_restarted)
ret_devices_up.extend(ret.get("devices_up", []))
failed_devices_up.extend(ret.get("failed_devices_up", []))
ret_devices_down.extend(ret.get("devices_down", []))
failed_devices_down.extend(ret.get("failed_devices_down", []))
return {'devices_up': ret_devices_up,
'failed_devices_up': failed_devices_up,
'devices_down': ret_devices_down,
'failed_devices_down': failed_devices_down}
def tunnel_sync(self, context, tunnel_ip, tunnel_type=None, host=None):
cctxt = self.client.prepare(version='1.4')

@ -259,3 +259,6 @@ VALUES_TYPE_RANGE = "range"
# for the restart success rate.
# [1] http://paste.openstack.org/show/745685/
AGENT_RES_PROCESSING_STEP = 100
# Number of resources for neutron to divide the large RPC
# call data sets.
RPC_RES_PROCESSING_STEP = 20

@ -317,9 +317,17 @@ class RpcCallbacksTestCase(base.BaseTestCase):
class RpcApiTestCase(base.BaseTestCase):
def _test_rpc_api(self, rpcapi, topic, method, rpc_method, **kwargs):
if method == "update_device_list":
expected = {'devices_up': [],
'failed_devices_up': [],
'devices_down': [],
'failed_devices_down': []}
else:
expected = 'foo'
ctxt = oslo_context.RequestContext(user='fake_user',
tenant='fake_project')
expected_retval = 'foo' if rpc_method == 'call' else None
expected_retval = expected if rpc_method == 'call' else None
expected_version = kwargs.pop('version', None)
fanout = kwargs.pop('fanout', False)

Loading…
Cancel
Save