rabittMQ 消息确认机制

http://www.rabbitmq.com/reliability.html

Reliability Guide

This page explains how to use the various features of AMQP and RabbitMQ to achieve reliable delivery - to ensure that messages are always delivered, even encountering failure in any part of your system.

What Can Fail?

Network problems are probably the most common class of failure. Not only can networks fail, firewalls can interrupt idle connections, and network failures are not always detected immediately.

In addition to connectivity failures, the broker and client applications can experience hardware failure (or software can crash) at any time. Additionally, even if client applications keep running, logic errors can cause channel or connection errors which force the client to establish a new channel or connection and recover from the problem.

Connection Failures

In the event of a connection failure, the client will need to establish a new connection to the broker. Any channels opened on the previous connection will have been automatically closed and these will need re-opening too.

In general when connections fail, the client will be informed by the connection throwing an exception (or similar language construct). The official Java and .NET clients additionally provide callback methods to let you hear about connection failures in other contexts - Java provides the ShutdownListener callback on both Connection and Channel classes, and .NET client provides IConnection.ConnectionShutdown andIModel.ModelShutdown events for the same purpose.

Acknowledgements and Confirms

When a connection fails, messages may be in transit between client and server - they may be in the middle of being parsed or generated, in OS buffers, or on the wire. Messages in transit will be lost - they will need to be retransmitted. Acknowledgements let the server and clients know when to do this.

Acknowledgements can be used in both directions - to allow a consumer to indicate to the server that it has received / processed a message and to allow the server to indicate the same thing to the producer. RabbitMQ refers to the latter case as a "confirm".

Of course, TCP ensures that packets have been received, and will retransmit until they are - but that's just the network layer. Acknowledgements and confirms indicate that messages have been received and acted upon. An acknowledgement signals both the receipt of a message, and a transfer of ownership where the receiver assumes full responsibility for it.

Acknowledgements therefore have semantics - a consuming application should not acknowledge messages until it has done whatever it needs to do with them - recorded them in a database, forwarded them on, printed them onto paper or anything else. Once it does so, the broker is free to forget about the message.

Similarly, the broker will confirm messages once it has taken responsibility for them (see here for what that means).

Use of acknowledgements guarantees at-least-once delivery. Without acknowledgements, message loss is possible during publish and consume operations and only at-most-once delivery is guaranteed.

Heartbeats

In some types of network failure, packet loss can mean that disrupted TCP connections take some time to be detected by the operating system. AMQP offers a heartbeat feature to ensure that the application layer promptly finds out about disrupted connections (and also completely unresponsive peers). Heartbeats also defend against certain network equipment which may terminate "idle" TCP connections. In RabbitMQ versions 3.0 and higher, the broker will attempt to negotiate heartbeats by default (although the client can still veto them). Using earlier versions the client must be configured to request heartbeats.

At the Broker

In order to avoid losing messages in the broker we need to cope with broker restarts, broker hardware failure and in extremis even broker crashes.

To ensure that messages and broker definitions survive restarts, we need to ensure that they are on disk. The AMQP standard has a concept of durability for exchanges, queues and of persistent messages, requiring that a durable object or persistent message will survive a restart. More details about specific flags pertaining to durability and persistence can be found in the AMQP Concepts Guide.

Clustering and High Availability

If we need to ensure that our broker survives hardware failure, we can use RabbitMQ's clustering. In a RabbitMQ cluster, all definitions (of exchanges, bindings, users, etc) are mirrored across the entire cluster. Queues behave differently, by default residing only on a single node, but optionally being mirrored across several or all nodes. Queues remain visible and reachable from all nodes regardless of where they are located.

Mirrored queues replicate their contents across all configured cluster nodes, tolerating node failures seamlessly and without message loss (although see this note on unsynchronised slaves). However, consuming applications need to be aware that when queues fail their consumers will be cancelled and they will need to reconsume - see the documentation for more details.

At the Producer

When using confirms, producers recovering from a channel or connection failure should retransmit any messages for which an acknowledgement has not been received from the broker. There is a possibility of message duplication here, because the broker might have sent a confirmation that never reached the producer (due to network failures, etc). Therefore consumer applications will need to perform deduplication or handle incoming messages in an idempotent manner.

Ensuring Messages are Routed

In some circumstances it can be important for producers to ensure that their messages are being routed to queues (although not always - in the case of a pub-sub system producers will just publish and if no consumers are interested it is correct for messages to be dropped).

To ensure messages are routed to a single known queue, the producer can just declare a destination queue and publish directly to it. If messages may be routed in more complex ways but the producer still needs to know if they reached at least one queue, it can set the mandatory flag on a basic.publish, ensuring that a basic.return (containing a reply code and some textual explanation) will be sent back to the client if no queues were appropriately bound.

Producers should also be aware that when publishing to a clustered node, if one or more destination queues that are bound to the exchange have mirrors in the cluster, it's possible to incur delays in the face of network failures between nodes, due to flow control between replicas and the master queue process. Seehere for more details.

At the Consumer

In the event of network failure (or a node crashing), messages can be duplicated, and consumers must be prepared to handle them. If possible, the simplest way to handle this is to ensure that your consumers handle messages in an idempotent way rather than explicitly deal with deduplication.

If a message is delivered to a consumer and then requeued (because it was not acknowledged before the consumer connection dropped, for example) then RabbitMQ will set the redelivered flag on it when it is delivered again (whether to the same consumer or a different one). This is a hint that a consumer may have seen this message before (although that's not guaranteed, the message may have made it out of the broker but not into a consumer before the connection dropped). Conversely if the redelivered flag is not set then it is guaranteed that the message has not been seen before. Therefore if a consumer finds it more expensive to deduplicate messages or process them in an idempotent manner, it can do this only for messages with theredelivered flag set.

Consumer Cancel Notification

Under some circumstances the server needs to be able to cancel a consumer - since the queue it was consuming from has been deleted, or has failed over. In this case the consumer should consume again but be aware that it may see messages again which it has already seen.

Note that consumer cancel notification is a RabbitMQ extension to AMQP, and as such may not be supported by all clients.

Messages That Cannot Be Processed

If a consumer determines that it cannot handle a message then it can reject it using basic.reject (orbasic.nack), either asking the server to requeue it, or not (in which case the server might be configured todead-letter it instead.

Distributed RabbitMQ

Rabbit provides two plugins to assist with distributing nodes over unreliable networks: federation and theshovel. Both are implemented as AMQP clients, so if you configure them to use confirms and acknowledgements, they will retransmit when necessary. Both will use confirms and acknowledgements by default.

When connecting clusters with federation or the shovel, it is desirable to ensure that the federation links and shovels tolerate node failures. Federation will automatically distribute links across the downstream cluster and fail them over on failure of a downstream node. In order to connect to a new upstream when an upstream node fails you can specify multiple redundant URIs for an upstream, or connect via a TCP load balancer.

When using the shovel, it is possible to specify redundant brokers in a source or destination clause; however it is not currently possible to make the shovel itself redundant. We hope to improve this situation in the future; in the mean time a new node can be brought up manually to run a shovel if the node it was originally running on fails.

 

以下是需要注意的方面,以python为例,原因一样,服务器有时内存占用大,答案也能在此找到.

（使用pika 0.9.5 Python客户端）

从哪里获得帮助

如果你在这篇教程中遇到了困难，你可以通过讨论列表或者直接 联系我们

工作队列（又称：任务队列——Task Queues）是为了避免等待一些占用大量资源、时间的操作。当我们把任务（Task）当作消息发送到队列中，一个运行在后台的工作者（worker）进程就会取出任务然后处理。当你运行多个工作者（workers），任务就会在它们之间共享。

这个概念在网络应用中是非常有用的，它可以在短暂的HTTP请求中处理一些复杂的任务。

准备

之前的教程中，我们发送了一个包含“Hello World!”的字符串消息。现在，我们将发送一些字符串，把这些字符串当作复杂的任务。我们没有真是的例子，例如图片缩放、pdf文件转换。所以使用time.sleep()函数来模拟这种情况。我们在字符串中加上点号（.）来表示任务的复杂程度，一个点（.）将会耗时1秒钟。比如"Hello..."就会耗时3秒钟。

我们对之前教程的send.py做些简单的调整，以便可以发送随意的消息。这个程序会按照计划发送任务到我们的工作队列中。我们把它命名为new_task.py：

import sys

message = ' '.join(sys.argv[1:]) or "Hello World!"
channel.basic_publish(exchange='',
                      routing_key='hello',
                      body=message)
print " [x] Sent %r" % (message,)

我们的旧脚本（receive.py）同样需要做一些改动：它需要为消息体中每一个点号（.）模拟1秒钟的操作。它会从队列中获取消息并执行，我们把它命名为worker.py：

import time

def callback(ch, method, properties, body):
    print " [x] Received %r" % (body,)
    time.sleep( body.count('.') )
    print " [x] Done"

轮询分发

使用工作队列的一个好处就是它能够并行的处理队列。如果堆积了很多任务，我们只需要添加更多的工作者（workers）就可以了，扩展很简单。

首先，我们先同时运行两个worker.py脚本，它们都会从队列中获取消息，到底是不是这样呢？我们看看。

你需要打开三个终端，两个用来运行worker.py脚本，这两个终端就是我们的两个消费者（consumers）—— C1 和 C2。

shell1$ python worker.py
 [*] Waiting for messages. To exit press CTRL+C

shell2$ python worker.py
 [*] Waiting for messages. To exit press CTRL+C

第三个终端，我们用来发布新任务。你可以发送一些消息给消费者（consumers）：

shell3$ python new_task.py First message.
shell3$ python new_task.py Second message..
shell3$ python new_task.py Third message...
shell3$ python new_task.py Fourth message....
shell3$ python new_task.py Fifth message.....

看看到底发送了什么给我们的工作者（workers）：

shell1$ python worker.py
 [*] Waiting for messages. To exit press CTRL+C
 [x] Received 'First message.'
 [x] Received 'Third message...'
 [x] Received 'Fifth message.....'

shell2$ python worker.py
 [*] Waiting for messages. To exit press CTRL+C
 [x] Received 'Second message..'
 [x] Received 'Fourth message....'

默认来说，RabbitMQ会按顺序得把消息发送给每个消费者（consumer）。平均每个消费者都会收到同等数量得消息。这种发送消息得方式叫做——轮询（round-robin）。试着添加三个或更多得工作者（workers）。

消息响应

当处理一个比较耗时得任务的时候，你也许想知道消费者（consumers）是否运行到一半就挂掉。当前的代码中，当消息被RabbitMQ发送给消费者（consumers）之后，马上就会在内存中移除。这种情况，你只要把一个工作者（worker）停止，正在处理的消息就会丢失。同时，所有发送到这个工作者的还没有处理的消息都会丢失。

我们不想丢失任何任务消息。如果一个工作者（worker）挂掉了，我们希望任务会重新发送给其他的工作者（worker）。

为了防止消息丢失，RabbitMQ提供了消息响应（acknowledgments）。消费者会通过一个ack（响应），告诉RabbitMQ已经收到并处理了某条消息，然后RabbitMQ就会释放并删除这条消息。

如果消费者（consumer）挂掉了，没有发送响应，RabbitMQ就会认为消息没有被完全处理，然后重新发送给其他消费者（consumer）。这样，及时工作者（workers）偶尔的挂掉，也不会丢失消息。

消息是没有超时这个概念的；当工作者与它断开连的时候，RabbitMQ会重新发送消息。这样在处理一个耗时非常长的消息任务的时候就不会出问题了。

消息响应默认是开启的。之前的例子中我们可以使用no_ack=True标识把它关闭。是时候移除这个标识了，当工作者（worker）完成了任务，就发送一个响应。

def callback(ch, method, properties, body):
    print " [x] Received %r" % (body,)
    time.sleep( body.count('.') )
    print " [x] Done"
    ch.basic_ack(delivery_tag = method.delivery_tag)

channel.basic_consume(callback,
                      queue='hello')

运行上面的代码，我们发现即使使用CTRL+C杀掉了一个工作者（worker）进程，消息也不会丢失。当工作者（worker）挂掉这后，所有没有响应的消息都会重新发送。

忘了响应

一个很容易犯的错误就是忘了basic_ack，后果很严重。消息在你的程序退出之后就会重新发送，如果它不能够释放没响应的消息，RabbitMQ就会占用越来越多的内存。

为了排除这种错误，你可以使用rabbitmqctl命令，输出messages_unacknowledged字段：

$ sudo rabbitmqctl list_queues name messages_ready messages_unacknowledged
Listing queues ...
hello    0       0
...done.

消息持久化

如果你没有特意告诉RabbitMQ，那么在它退出或者崩溃的时候，它将会流失所有的队列和消息。为了确保信息不会丢失，有两个事情是需要注意的：我们必须把“队列”和“消息”设为持久化。

首先，为了不让队列丢失，需要把它声明为持久化（durable）：

channel.queue_declare(queue='hello', durable=True)

尽管这行代码本身是正确的，但是仍然不会正确运行。因为我们已经定义过一个叫hello的非持久化队列。RabbitMq不允许你使用不同的参数重新定义一个队列，它会返回一个错误。但我们现在使用一个快捷的解决方法——用不同的名字，例如task_queue。

channel.queue_declare(queue='task_queue', durable=True)

这个queue_declare必须在生产者（producer）和消费者（consumer）对应的代码中修改。

这时候，我们就可以确保在RabbitMq重启之后queue_declare队列不会丢失。另外，我们需要把我们的消息也要设为持久化——将delivery_mode的属性设为2。

channel.basic_publish(exchange='',
                      routing_key="task_queue",
                      body=message,
                      properties=pika.BasicProperties(
                         delivery_mode = 2, # make message persistent
                      ))

注意：消息持久化

将消息设为持久化并不能完全保证不会丢失。以上代码只是告诉了RabbitMq要把消息存到硬盘，但从RabbitMq收到消息到保存之间还是有一个很小的间隔时间。因为RabbitMq并不是所有的消息都使用fsync(2)——它有可能只是保存到缓存中，并不一定会写到硬盘中。并不能保证真正的持久化，但已经足够应付我们的简单工作队列。如果你一定要保证持久化，你需要改写你的代码来支持事务（transaction）。

公平分发

你应该已经发现，它仍旧没有按照我们期望的那样进行分发。比如有两个工作者（workers），处理奇数消息的比较繁忙，处理偶数消息的比较轻松。然而RabbitMQ并不知道这些，它仍然一如既往的派发消息。

这时因为RabbitMQ只管分发进入队列的消息，不会关心有多少消费者（consumer）没有作出响应。它盲目的把第n-th条消息发给第n-th个消费者。

我们可以使用basic.qos方法，并设置prefetch_count=1。这样是告诉RabbitMQ，再同一时刻，不要发送超过1条消息给一个工作者（worker），直到它已经处理了上一条消息并且作出了响应。这样，RabbitMQ就会把消息分发给下一个空闲的工作者（worker）。

channel.basic_qos(prefetch_count=1)

关于队列大小

如果所有的工作者都处理繁忙状态，你的队列就会被填满。你需要留意这个问题，要么添加更多的工作者（workers），要么使用其他策略。

整合

new_task.py的完整代码：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

#!/usr/bin/env python import pika import sys connection = pika.BlockingConnection(pika.ConnectionParameters( host='localhost')) channel = connection.channel() channel.queue_declare(queue='task_queue', durable=True) message = ' '.join(sys.argv[1:]) or "Hello World!" channel.basic_publish(exchange='', routing_key='task_queue', body=message, properties=pika.BasicProperties( delivery_mode = 2, # make message persistent )) print " [x] Sent %r" % (message,) connection.close()

(new_task.py源码)

我们的worker：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

#!/usr/bin/env python import pika import time connection = pika.BlockingConnection(pika.ConnectionParameters( host='localhost')) channel = connection.channel() channel.queue_declare(queue='task_queue', durable=True) print ' [*] Waiting for messages. To exit press CTRL+C' def callback(ch, method, properties, body): print " [x] Received %r" % (body,) time.sleep( body.count('.') ) print " [x] Done" ch.basic_ack(delivery_tag = method.delivery_tag) channel.basic_qos(prefetch_count=1) channel.basic_consume(callback, queue='task_queue') channel.start_consuming()

(worker.py source)

使用消息响应和prefetch_count你就可以搭建起一个工作队列了。这些持久化的选项使得在RabbitMQ重启之后仍然能够恢复。

现在我们可以移步教程3学习如何发送相同的消息给多个消费者（consumers）。