oslo.messaging/oslo_messaging/rpc/dispatcher.py
Arnaud Morin 82492442f3 Add a ping endpoint to RPC dispatcher
The purpose of this patch is to add an endpoint directly in RPC
dispatcher, so this endpoint will always be available, in a cross
project manner, without the need for projects to manage it by themself.

This endpoint stay disabled by default, so this change is harmless
without a specific configuration option.

To enable this ping endpoint, an operator will just have to add a new
parameter in the [DEFAULT] section, alongside with rpc_response_timeout
[DEFAULT]
rpc_ping_enabled=true  # default is false

The purpose of this new endpoint is to help operators do a RPC call (a
ping) toward a specific RPC callback (e.g. a nova-compute, or a
neutron-agent).
This is helping a lot for monitoring agents (for example, if agents are
deployed in a kubernetes pod).

The endpoint is named oslo_rpc_server_ping.

Change-Id: I51cf67e060f240e6eb82260e70a057fe599f9063
Signed-off-by: Arnaud Morin <arnaud.morin@corp.ovh.com>
2020-08-18 15:09:29 +02:00

321 lines
12 KiB
Python

# Copyright 2010 United States Government as represented by the
# Administrator of the National Aeronautics and Space Administration.
# All Rights Reserved.
# Copyright 2013 Red Hat, Inc.
# Copyright 2013 New Dream Network, LLC (DreamHost)
#
# 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.
from abc import ABCMeta
from abc import abstractmethod
import logging
import sys
import threading
from oslo_config import cfg
from oslo_utils import eventletutils
from oslo_messaging import _utils as utils
from oslo_messaging import dispatcher
from oslo_messaging import serializer as msg_serializer
from oslo_messaging import server as msg_server
from oslo_messaging import target as msg_target
_dispatcher_opts = [
cfg.BoolOpt('rpc_ping_enabled',
default=False,
help='Add an endpoint to answer to ping calls. '
'Endpoint is named oslo_rpc_server_ping'),
]
__all__ = [
'NoSuchMethod',
'RPCAccessPolicyBase',
'LegacyRPCAccessPolicy',
'DefaultRPCAccessPolicy',
'ExplicitRPCAccessPolicy',
'RPCDispatcher',
'RPCDispatcherError',
'UnsupportedVersion',
'ExpectedException',
]
LOG = logging.getLogger(__name__)
class PingEndpoint(object):
def oslo_rpc_server_ping(self, ctxt, **kwargs):
return 'pong'
class ExpectedException(Exception):
"""Encapsulates an expected exception raised by an RPC endpoint
Merely instantiating this exception records the current exception
information, which will be passed back to the RPC client without
exceptional logging.
"""
def __init__(self):
self.exc_info = sys.exc_info()
class RPCDispatcherError(msg_server.MessagingServerError):
"A base class for all RPC dispatcher exceptions."
class NoSuchMethod(RPCDispatcherError, AttributeError):
"Raised if there is no endpoint which exposes the requested method."
def __init__(self, method):
msg = "Endpoint does not support RPC method %s" % method
super(NoSuchMethod, self).__init__(msg)
self.method = method
class UnsupportedVersion(RPCDispatcherError):
"Raised if there is no endpoint which supports the requested version."
def __init__(self, version, method=None):
msg = "Endpoint does not support RPC version %s" % version
if method:
msg = "%s. Attempted method: %s" % (msg, method)
super(UnsupportedVersion, self).__init__(msg)
self.version = version
self.method = method
class RPCAccessPolicyBase(object, metaclass=ABCMeta):
"""Determines which endpoint methods may be invoked via RPC"""
@abstractmethod
def is_allowed(self, endpoint, method):
"""Applies an access policy to the rpc method
:param endpoint: the instance of a rpc endpoint
:param method: the method of the endpoint
:return: True if the method may be invoked via RPC, else False.
"""
class LegacyRPCAccessPolicy(RPCAccessPolicyBase):
"""The legacy access policy allows RPC access to all callable endpoint
methods including private methods (methods prefixed by '_')
"""
def is_allowed(self, endpoint, method):
return True
class DefaultRPCAccessPolicy(RPCAccessPolicyBase):
"""The default access policy prevents RPC calls to private methods
(methods prefixed by '_')
.. note::
LegacyRPCAdapterPolicy currently needs to be the default while we have
projects that rely on exposing private methods.
"""
def is_allowed(self, endpoint, method):
return not method.startswith('_')
class ExplicitRPCAccessPolicy(RPCAccessPolicyBase):
"""Policy which requires decorated endpoint methods to allow dispatch"""
def is_allowed(self, endpoint, method):
if hasattr(endpoint, method):
return hasattr(getattr(endpoint, method), 'exposed')
return False
class RPCDispatcher(dispatcher.DispatcherBase):
"""A message dispatcher which understands RPC messages.
A MessageHandlingServer is constructed by passing a callable dispatcher
which is invoked with context and message dictionaries each time a message
is received.
RPCDispatcher is one such dispatcher which understands the format of RPC
messages. The dispatcher looks at the namespace, version and method values
in the message and matches those against a list of available endpoints.
Endpoints may have a target attribute describing the namespace and version
of the methods exposed by that object.
The RPCDispatcher may have an access_policy attribute which determines
which of the endpoint methods are to be dispatched.
The default access_policy dispatches all public methods
on an endpoint object.
"""
def __init__(self, endpoints, serializer, access_policy=None):
"""Construct a rpc server dispatcher.
:param endpoints: list of endpoint objects for dispatching to
:param serializer: optional message serializer
"""
cfg.CONF.register_opts(_dispatcher_opts)
oslo_rpc_server_ping = None
for ep in endpoints:
# Check if we have an attribute named 'target'
target = getattr(ep, 'target', None)
if target and not isinstance(target, msg_target.Target):
errmsg = "'target' is a reserved Endpoint attribute used" + \
" for namespace and version filtering. It must" + \
" be of type oslo_messaging.Target. Do not" + \
" define an Endpoint method named 'target'"
raise TypeError("%s: endpoint=%s" % (errmsg, ep))
# Check if we have an attribute named 'oslo_rpc_server_ping'
oslo_rpc_server_ping = getattr(ep, 'oslo_rpc_server_ping', None)
if oslo_rpc_server_ping:
errmsg = "'oslo_rpc_server_ping' is a reserved Endpoint" + \
" attribute which can be use to ping the" + \
" endpoint. Please avoid using any oslo_* " + \
" naming."
LOG.warning("%s (endpoint=%s)" % (errmsg, ep))
self.endpoints = endpoints
# Add ping endpoint if enabled in config
if cfg.CONF.rpc_ping_enabled:
if oslo_rpc_server_ping:
LOG.warning("rpc_ping_enabled=True in config but "
"oslo_rpc_server_ping is already declared "
"in an other Endpoint. Not enabling rpc_ping "
"Endpoint.")
else:
self.endpoints.append(PingEndpoint())
self.serializer = serializer or msg_serializer.NoOpSerializer()
self._default_target = msg_target.Target()
if access_policy is not None:
if issubclass(access_policy, RPCAccessPolicyBase):
self.access_policy = access_policy()
else:
raise TypeError('access_policy must be a subclass of '
'RPCAccessPolicyBase')
else:
self.access_policy = DefaultRPCAccessPolicy()
@staticmethod
def _is_namespace(target, namespace):
return namespace in target.accepted_namespaces
@staticmethod
def _is_compatible(target, version):
endpoint_version = target.version or '1.0'
return utils.version_is_compatible(endpoint_version, version)
def _do_dispatch(self, endpoint, method, ctxt, args):
ctxt = self.serializer.deserialize_context(ctxt)
new_args = dict()
for argname, arg in args.items():
new_args[argname] = self.serializer.deserialize_entity(ctxt, arg)
func = getattr(endpoint, method)
result = func(ctxt, **new_args)
return self.serializer.serialize_entity(ctxt, result)
def _watchdog(self, event, incoming):
# NOTE(danms): If the client declared that they are going to
# time out after N seconds, send the call-monitor heartbeat
# every N/2 seconds to make sure there is plenty of time to
# account for inbound and outbound queuing delays. Client
# timeouts must be integral and positive, otherwise we log and
# ignore.
try:
client_timeout = int(incoming.client_timeout)
cm_heartbeat_interval = client_timeout / 2
except ValueError:
client_timeout = cm_heartbeat_interval = 0
if cm_heartbeat_interval < 1:
LOG.warning('Client provided an invalid timeout value of %r' % (
incoming.client_timeout))
return
while not event.wait(cm_heartbeat_interval):
LOG.debug(
'Sending call-monitor heartbeat for active call to %(method)s '
'(interval=%(interval)i)' % (
{'method': incoming.message.get('method'),
'interval': cm_heartbeat_interval}))
try:
incoming.heartbeat()
except Exception as exc:
# The heartbeat message failed to send. Likely the broker or
# client has died. Nothing to do here but exit the watchdog
# thread. If the client is still alive (dead broker) then its
# RPC will timeout as expected.
LOG.debug("Call-monitor heartbeat failed: %(exc)s"
% ({'exc': exc}))
break
def dispatch(self, incoming):
"""Dispatch an RPC message to the appropriate endpoint method.
:param incoming: incoming message
:type incoming: IncomingMessage
:raises: NoSuchMethod, UnsupportedVersion
"""
message = incoming.message
ctxt = incoming.ctxt
method = message.get('method')
args = message.get('args', {})
namespace = message.get('namespace')
version = message.get('version', '1.0')
# NOTE(danms): This event and watchdog thread are used to send
# call-monitoring heartbeats for this message while the call
# is executing if it runs for some time. The thread will wait
# for the event to be signaled, which we do explicitly below
# after dispatching the method call.
completion_event = eventletutils.Event()
watchdog_thread = threading.Thread(target=self._watchdog,
args=(completion_event, incoming))
if incoming.client_timeout:
# NOTE(danms): The client provided a timeout, so we start
# the watchdog thread. If the client is old or didn't send
# a timeout, we just never start the watchdog thread.
watchdog_thread.start()
found_compatible = False
for endpoint in self.endpoints:
target = getattr(endpoint, 'target', None)
if not target:
target = self._default_target
if not (self._is_namespace(target, namespace) and
self._is_compatible(target, version)):
continue
if hasattr(endpoint, method):
if self.access_policy.is_allowed(endpoint, method):
try:
return self._do_dispatch(endpoint, method, ctxt, args)
finally:
completion_event.set()
if incoming.client_timeout:
watchdog_thread.join()
found_compatible = True
if found_compatible:
raise NoSuchMethod(method)
else:
raise UnsupportedVersion(version, method=method)