openstack/oslo.messaging
Code Issues Proposed changes
oslo.messaging/oslo_messaging/_drivers/protocols/amqp/controller.py
Kenneth Giusti c127594de6 Provide better detection of failures during message send 
This change causes the message sender to block until the messaging
infrastructure (e.g. the broker) assumes ownership of the message (or
fails to accept it).  If the message is accepted, then the sender will
either pend for a response (in the case of RPC), or simply return (in
the case of notification).  If the message is rejected by the
messaging infrastructure a MessagingException will be raised at the
sender.

Change-Id: I3f4a1ed1c17e18f6d629f16e6b5c99de45b083d6
Closes-Bug: #1377228
2015-06-08 20:42:10 -04:00

630 lines
26 KiB
Python
Raw Blame History

# Copyright 2014, Red Hat, Inc.
#
# 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.
"""
Controller that manages the interface between the driver and the messaging
service.
This module defines a Controller class that is responsible for performing
messaging-related operations (Tasks) requested by the driver, and for managing
the connection to the messaging service. The Controller creates a background
thread which performs all messaging operations and socket I/O. The
Controller's messaging logic is executed in the background thread via lambda
functions scheduled by the Controller.
"""
import abc
import logging
import threading
import uuid
from oslo_config import cfg
import proton
import pyngus
from six import moves
from oslo_messaging._drivers.protocols.amqp import eventloop
from oslo_messaging._drivers.protocols.amqp import opts
from oslo_messaging import exceptions
from oslo_messaging import transport
LOG = logging.getLogger(__name__)
class Task(object):
"""Perform a messaging operation via the Controller."""
@abc.abstractmethod
def execute(self, controller):
"""This method will be run on the eventloop thread."""
class Replies(pyngus.ReceiverEventHandler):
"""This is the receiving link for all reply messages. Messages are routed
to the proper Listener's incoming queue using the correlation-id header in
the message.
"""
def __init__(self, connection, on_ready):
self._correlation = {} # map of correlation-id to response queue
self._ready = False
self._on_ready = on_ready
rname = "Consumer-%s:src=[dynamic]:tgt=replies" % uuid.uuid4().hex
self._receiver = connection.create_receiver("replies",
event_handler=self,
name=rname)
# capacity determines the maximum number of reply messages this link
# can receive. As messages are received and credit is consumed, this
# driver will 'top up' the credit back to max capacity. This number
# should be large enough to avoid needlessly flow-controlling the
# replies.
self.capacity = 100 # TODO(kgiusti) guesstimate - make configurable
self._credit = 0
self._receiver.open()
def ready(self):
return self._ready
def prepare_for_response(self, request, reply_queue):
"""Apply a unique message identifier to this request message. This will
be used to identify messages sent in reply. The identifier is placed
in the 'id' field of the request message. It is expected that the
identifier will appear in the 'correlation-id' field of the
corresponding response message.
"""
request.id = uuid.uuid4().hex
# reply is placed on reply_queue
self._correlation[request.id] = reply_queue
request.reply_to = self._receiver.source_address
LOG.debug("Reply for msg id=%s expected on link %s",
request.id, request.reply_to)
return request.id
def cancel_response(self, msg_id):
"""Abort waiting for a response message. This can be used if the
request fails and no reply is expected.
"""
if msg_id in self._correlation:
del self._correlation[msg_id]
# Pyngus ReceiverLink event callbacks:
def receiver_active(self, receiver_link):
"""This is a Pyngus callback, invoked by Pyngus when the receiver_link
has transitioned to the open state and is able to receive incoming
messages.
"""
self._ready = True
self._update_credit()
self._on_ready()
LOG.debug("Replies expected on link %s",
self._receiver.source_address)
def receiver_remote_closed(self, receiver, pn_condition):
"""This is a Pyngus callback, invoked by Pyngus when the peer of this
receiver link has initiated closing the connection.
"""
# TODO(kgiusti) Unclear if this error will ever occur (as opposed to
# the Connection failing instead). Log for now, possibly implement a
# recovery strategy if necessary.
LOG.error("Reply subscription closed by peer: %s",
(pn_condition or "no error given"))
def message_received(self, receiver, message, handle):
"""This is a Pyngus callback, invoked by Pyngus when a new message
arrives on this receiver link from the peer.
"""
self._credit = self._credit - 1
self._update_credit()
key = message.correlation_id
if key in self._correlation:
LOG.debug("Received response for msg id=%s", key)
result = {"status": "OK",
"response": message}
self._correlation[key].put(result)
# cleanup (only need one response per request)
del self._correlation[key]
receiver.message_accepted(handle)
else:
LOG.warn("Can't find receiver for response msg id=%s, dropping!",
key)
receiver.message_modified(handle, True, True, None)
def _update_credit(self):
# ensure we have enough credit
if self._credit < self.capacity / 2:
self._receiver.add_capacity(self.capacity - self._credit)
self._credit = self.capacity
class Server(pyngus.ReceiverEventHandler):
"""A group of links that receive messages from a set of addresses derived
from a given target. Messages arriving on the links are placed on the
'incoming' queue.
"""
def __init__(self, addresses, incoming):
self._incoming = incoming
self._addresses = addresses
self._capacity = 500 # credit per link
def attach(self, connection):
"""Create receiver links over the given connection for all the
configured addresses.
"""
self._receivers = []
for a in self._addresses:
props = {"snd-settle-mode": "settled"}
rname = "Consumer-%s:src=%s:tgt=%s" % (uuid.uuid4().hex, a, a)
r = connection.create_receiver(source_address=a,
target_address=a,
event_handler=self,
name=rname,
properties=props)
# TODO(kgiusti) Hardcoding credit here is sub-optimal. A better
# approach would monitor for a back-up of inbound messages to be
# processed by the consuming application and backpressure the
# sender based on configured thresholds.
r.add_capacity(self._capacity)
r.open()
self._receivers.append(r)
# Pyngus ReceiverLink event callbacks:
def receiver_remote_closed(self, receiver, pn_condition):
"""This is a Pyngus callback, invoked by Pyngus when the peer of this
receiver link has initiated closing the connection.
"""
text = "Server subscription %(addr)s closed by peer: %(err_msg)s"
vals = {
"addr": receiver.source_address or receiver.target_address,
"err_msg": pn_condition or "no error given"
}
LOG.error(text % vals)
def message_received(self, receiver, message, handle):
"""This is a Pyngus callback, invoked by Pyngus when a new message
arrives on this receiver link from the peer.
"""
if receiver.capacity < self._capacity / 2:
receiver.add_capacity(self._capacity - receiver.capacity)
self._incoming.put(message)
LOG.debug("message received: %s", message)
receiver.message_accepted(handle)
class Hosts(object):
"""An order list of TransportHost addresses. Connection failover
progresses from one host to the next.
"""
def __init__(self, entries=None):
self._entries = entries[:] if entries else []
for entry in self._entries:
entry.port = entry.port or 5672
self._current = 0
def add(self, transport_host):
self._entries.append(transport_host)
@property
def current(self):
if len(self._entries):
return self._entries[self._current]
else:
return transport.TransportHost(hostname="localhost", port=5672)
def next(self):
if len(self._entries) > 1:
self._current = (self._current + 1) % len(self._entries)
return self.current
def __repr__(self):
return '<Hosts ' + str(self) + '>'
def __str__(self):
return ", ".join(["%r" % th for th in self._entries])
class Controller(pyngus.ConnectionEventHandler):
"""Controls the connection to the AMQP messaging service. This object is
the 'brains' of the driver. It maintains the logic for addressing, sending
and receiving messages, and managing the connection. All messaging and I/O
work is done on the Eventloop thread, allowing the driver to run
asynchronously from the messaging clients.
"""
def __init__(self, hosts, default_exchange, config):
self.processor = None
# queue of Task() objects to execute on the eventloop once the
# connection is ready:
self._tasks = moves.queue.Queue(maxsize=500)
# limit the number of Task()'s to execute per call to _process_tasks().
# This allows the eventloop main thread to return to servicing socket
# I/O in a timely manner
self._max_task_batch = 50
# cache of sending links indexed by address:
self._senders = {}
# Servers (set of receiving links), indexed by target:
self._servers = {}
self.hosts = Hosts(hosts)
opt_group = cfg.OptGroup(name='oslo_messaging_amqp',
title='AMQP 1.0 driver options')
config.register_group(opt_group)
config.register_opts(opts.amqp1_opts, group=opt_group)
self.server_request_prefix = \
config.oslo_messaging_amqp.server_request_prefix
self.broadcast_prefix = config.oslo_messaging_amqp.broadcast_prefix
self.group_request_prefix = \
config.oslo_messaging_amqp.group_request_prefix
self._container_name = config.oslo_messaging_amqp.container_name
self.idle_timeout = config.oslo_messaging_amqp.idle_timeout
self.trace_protocol = config.oslo_messaging_amqp.trace
self.ssl_ca_file = config.oslo_messaging_amqp.ssl_ca_file
self.ssl_cert_file = config.oslo_messaging_amqp.ssl_cert_file
self.ssl_key_file = config.oslo_messaging_amqp.ssl_key_file
self.ssl_key_password = config.oslo_messaging_amqp.ssl_key_password
self.ssl_allow_insecure = \
config.oslo_messaging_amqp.allow_insecure_clients
self.separator = "."
self.fanout_qualifier = "all"
self.default_exchange = default_exchange
# can't handle a request until the replies link is active, as
# we need the peer assigned address, so need to delay any
# processing of task queue until this is done
self._replies = None
# Set True when the driver is shutting down
self._closing = False
# only schedule one outstanding reconnect attempt at a time
self._reconnecting = False
self._delay = 0 # seconds between retries
# prevent queuing up multiple requests to run _process_tasks()
self._process_tasks_scheduled = False
self._process_tasks_lock = threading.Lock()
def connect(self):
"""Connect to the messaging service."""
self.processor = eventloop.Thread(self._container_name)
self.processor.wakeup(lambda: self._do_connect())
def add_task(self, task):
"""Add a Task for execution on processor thread."""
self._tasks.put(task)
self._schedule_task_processing()
def shutdown(self, wait=True, timeout=None):
"""Shutdown the messaging service."""
if self.processor:
LOG.debug("Waiting for eventloop to exit")
self.processor.shutdown(wait, timeout)
self.processor = None
LOG.debug("Eventloop exited, driver shut down")
# The remaining methods are reserved to run from the eventloop thread only!
# They must not be invoked directly!
# methods executed by Tasks created by the driver:
def request(self, target, request, result_queue, reply_expected=False):
"""Send a request message to the given target and arrange for a
result to be put on the result_queue. If reply_expected, the result
will include the reply message (if successful).
"""
address = self._resolve(target)
LOG.debug("Sending request for %s to %s", target, address)
if reply_expected:
msg_id = self._replies.prepare_for_response(request, result_queue)
def _callback(link, handle, state, info):
if state == pyngus.SenderLink.ACCEPTED: # message received
if not reply_expected:
# can wake up the sender now
result = {"status": "OK"}
result_queue.put(result)
else:
# we will wake up the sender when the reply message is
# received. See Replies.message_received()
pass
else: # send failed/rejected/etc
msg = "Message send failed: remote disposition: %s, info: %s"
exc = exceptions.MessageDeliveryFailure(msg % (state, info))
result = {"status": "ERROR", "error": exc}
if reply_expected:
# no response will be received, so cancel the correlation
self._replies.cancel_response(msg_id)
result_queue.put(result)
self._send(address, request, _callback)
def response(self, address, response):
"""Send a response message to the client listening on 'address'.
To prevent a misbehaving client from blocking a server indefinitely,
the message is send asynchronously.
"""
LOG.debug("Sending response to %s", address)
self._send(address, response)
def subscribe(self, target, in_queue):
"""Subscribe to messages sent to 'target', place received messages on
'in_queue'.
"""
addresses = [
self._server_address(target),
self._broadcast_address(target),
self._group_request_address(target)
]
self._subscribe(target, addresses, in_queue)
def subscribe_notifications(self, target, in_queue):
"""Subscribe for notifications on 'target', place received messages on
'in_queue'.
"""
addresses = [self._group_request_address(target)]
self._subscribe(target, addresses, in_queue)
def _subscribe(self, target, addresses, in_queue):
LOG.debug("Subscribing to %s (%s)", target, addresses)
self._servers[target] = Server(addresses, in_queue)
self._servers[target].attach(self._socket_connection.connection)
def _resolve(self, target):
"""Return a link address for a given target."""
if target.server:
return self._server_address(target)
elif target.fanout:
return self._broadcast_address(target)
else:
return self._group_request_address(target)
def _sender(self, address):
# if we already have a sender for that address, use it
# else establish the sender and cache it
if address in self._senders:
sender = self._senders[address]
else:
sname = "Producer-%s:src=%s:tgt=%s" % (uuid.uuid4().hex,
address, address)
conn = self._socket_connection.connection
sender = conn.create_sender(source_address=address,
target_address=address,
name=sname)
sender.open()
self._senders[address] = sender
return sender
def _send(self, addr, message, callback=None, handle=None):
"""Send the message out the link addressed by 'addr'. If a
delivery_callback is given it will be invoked when the send has
completed (whether successfully or in error).
"""
address = str(addr)
message.address = address
self._sender(address).send(message, delivery_callback=callback)
def _server_address(self, target):
return self._concatenate([self.server_request_prefix,
target.exchange or self.default_exchange,
target.topic, target.server])
def _broadcast_address(self, target):
return self._concatenate([self.broadcast_prefix,
target.exchange or self.default_exchange,
target.topic, self.fanout_qualifier])
def _group_request_address(self, target):
return self._concatenate([self.group_request_prefix,
target.exchange or self.default_exchange,
target.topic])
def _concatenate(self, items):
return self.separator.join(filter(bool, items))
# commands executed on the processor (eventloop) via 'wakeup()':
def _do_connect(self):
"""Establish connection and reply subscription on processor thread."""
host = self.hosts.current
conn_props = {}
if self.idle_timeout:
conn_props["idle-time-out"] = float(self.idle_timeout)
if self.trace_protocol:
conn_props["x-trace-protocol"] = self.trace_protocol
if self.ssl_ca_file:
conn_props["x-ssl-ca-file"] = self.ssl_ca_file
if self.ssl_cert_file:
# assume this connection is for a server. If client authentication
# support is developed, we'll need an explict flag (server or
# client)
conn_props["x-ssl-server"] = True
conn_props["x-ssl-identity"] = (self.ssl_cert_file,
self.ssl_key_file,
self.ssl_key_password)
conn_props["x-ssl-allow-cleartext"] = self.ssl_allow_insecure
self._socket_connection = self.processor.connect(host,
handler=self,
properties=conn_props)
LOG.debug("Connection initiated")
def _process_tasks(self):
"""Execute Task objects in the context of the processor thread."""
with self._process_tasks_lock:
self._process_tasks_scheduled = False
count = 0
while (not self._tasks.empty() and
count < self._max_task_batch and
self._can_process_tasks):
try:
self._tasks.get(False).execute(self)
except Exception as e:
LOG.exception("Error processing task: %s", e)
count += 1
# if we hit _max_task_batch, resume task processing later:
if not self._tasks.empty() and self._can_process_tasks:
self._schedule_task_processing()
def _schedule_task_processing(self):
"""_process_tasks() helper: prevent queuing up multiple requests for
task processing. This method is called both by the application thread
and the processing thread.
"""
if self.processor:
with self._process_tasks_lock:
already_scheduled = self._process_tasks_scheduled
self._process_tasks_scheduled = True
if not already_scheduled:
self.processor.wakeup(lambda: self._process_tasks())
@property
def _can_process_tasks(self):
"""_process_tasks helper(): indicates that the driver is ready to
process Tasks. In order to process messaging-related tasks, the reply
queue link must be active.
"""
return (not self._closing and
self._replies and self._replies.ready())
def _start_shutdown(self):
"""Called when the driver destroys the controller, this method attempts
to cleanly close the AMQP connection to the peer.
"""
LOG.info("Shutting down AMQP connection")
self._closing = True
if self._socket_connection.connection.active:
# try a clean shutdown
self._socket_connection.connection.close()
else:
# don't wait for a close from the remote, may never happen
self._complete_shutdown()
# reply link active callback:
def _reply_link_ready(self):
"""Invoked when the Replies reply link has become active. At this
point, we are ready to send/receive messages (via Task processing).
"""
LOG.info("Messaging is active (%s:%i)", self.hosts.current.hostname,
self.hosts.current.port)
self._schedule_task_processing()
# callback from eventloop on socket error
def socket_error(self, error):
"""Called by eventloop when a socket error occurs."""
LOG.debug("Socket failure: %s", error)
self._handle_connection_loss()
# Pyngus connection event callbacks (and their helpers), all invoked from
# the eventloop thread:
def connection_failed(self, connection, error):
"""This is a Pyngus callback, invoked by Pyngus when a non-recoverable
error occurs on the connection.
"""
if connection is not self._socket_connection.connection:
# pyngus bug: ignore failure callback on destroyed connections
return
LOG.debug("AMQP Connection failure: %s", error)
self._handle_connection_loss()
def connection_active(self, connection):
"""This is a Pyngus callback, invoked by Pyngus when the connection to
the peer is up. At this point, the driver will activate all subscriber
links (server) and the reply link.
"""
LOG.debug("Connection active (%s:%i), subscribing...",
self.hosts.current.hostname, self.hosts.current.port)
for s in self._servers.itervalues():
s.attach(self._socket_connection.connection)
self._replies = Replies(self._socket_connection.connection,
lambda: self._reply_link_ready())
self._delay = 0
def connection_closed(self, connection):
"""This is a Pyngus callback, invoked by Pyngus when the connection has
cleanly closed. This occurs after the driver closes the connection
locally, and the peer has acknowledged the close. At this point, the
shutdown of the driver's connection is complete.
"""
LOG.debug("AMQP connection closed.")
# if the driver isn't being shutdown, failover and reconnect
self._handle_connection_loss()
def connection_remote_closed(self, connection, reason):
"""This is a Pyngus callback, invoked by Pyngus when the peer has
requested that the connection be closed.
"""
if not self._closing:
# The messaging service/broker is trying to shut down the
# connection. Acknowledge the close, and try to reconnect/failover
# later once the connection has closed (connection_closed is
# called).
LOG.info("Connection closed by peer: %s",
reason or "no reason given")
self._socket_connection.connection.close()
def sasl_done(self, connection, pn_sasl, outcome):
"""This is a Pyngus callback invoked by Pyngus when the SASL handshake
has completed. The outcome of the handshake will be OK on success or
AUTH on failure.
"""
if outcome == proton.SASL.AUTH:
LOG.error("Unable to connect to %s:%s, authentication failure.",
self.hosts.current.hostname, self.hosts.current.port)
# requires user intervention, treat it like a connection failure:
self._handle_connection_loss()
def _complete_shutdown(self):
"""The AMQP Connection has closed, and the driver shutdown is complete.
Clean up controller resources and exit.
"""
self._socket_connection.close()
self.processor.shutdown()
LOG.info("Messaging has shutdown")
def _handle_connection_loss(self):
"""The connection to the messaging service has been lost. Try to
reestablish the connection/failover.
"""
if self._closing:
# we're in the middle of shutting down the driver anyways,
# just consider it done:
self._complete_shutdown()
else:
# for some reason, we've lost the connection to the messaging
# service. Try to re-establish the connection:
if not self._reconnecting:
self._reconnecting = True
self._replies = None
if self._delay == 0:
self._delay = 1
self._do_reconnect()
else:
d = self._delay
LOG.info("delaying reconnect attempt for %d seconds", d)
self.processor.schedule(lambda: self._do_reconnect(), d)
self._delay = min(d * 2, 60)
def _do_reconnect(self):
"""Invoked on connection/socket failure, failover and re-connect to the
messaging service.
"""
if not self._closing:
self._reconnecting = False
self._senders = {}
self._socket_connection.reset()
host = self.hosts.next()
LOG.info("Reconnecting to: %s:%i", host.hostname, host.port)
self._socket_connection.connect(host)
View Git Blame Copy Permalink