RabbitMQ Work Queue模式讲解

RabbitMQ Work Queue模式讲解

一．W/Q 模式的结构

有一个负责发送任务，多个执行单元接受任务后进行处理。该模式中，没个消费者接受的任务量是均衡的。不会根据具体任务的处理时间进行更精确的分配

发布者代码
var message = GetMessage(args);var body = Encoding.UTF8.GetBytes(message);
var properties = channel.CreateBasicProperties();
properties.Persistent = true;

channel.BasicPublish(exchange: “”,
routingKey: “task_queue”,
basicProperties: properties,
body: body);

接收者代码
var consumer = new EventingBasicConsumer(channel);
consumer.Received += (model, ea) =>
{
var body = ea.Body.ToArray();
var message = Encoding.UTF8.GetString(body);
Console.WriteLine(" [x] Received {0}", message);

int dots = message.Split('.').Length - 1;
Thread.Sleep(dots * 1000);

Console.WriteLine(" [x] Done");

};
channel.BasicConsume(queue: “task_queue”, autoAck: true, consumer: consumer);

发送者的Console

shell 3cd NewTask

dotnet run “First message.”
dotnet run “Second message…”
dotnet run “Third message…”
dotnet run “Fourth message…”
dotnet run “Fifth message…”

Worker1 的Console

shell 1# => [*] Waiting for messages. To exit press CTRL+C# => [x] Received ‘First message.’# => [x] Received ‘Third message…’# => [x] Received ‘Fifth message…’

Work2的Console

shell 2# => [*] Waiting for messages. To exit press CTRL+C# => [x] Received ‘Second message…’# => [x] Received ‘Fourth message…’

以上的，在两个接受者收到消息后，消息队列就自己删除了，不会等待接受者对于任务处理的反馈。

二.消息的确认

如果我们不想在Woker失效（通道关闭，连接关闭，TCP断开）的时候，导致消息被误删除，我们就要对message acknowlegements进行修改，当前Woker不能完成的时候，MQ会对该消息进行转发，到其他工作的Worker上进行处理。

Worker的代码 ，修改 AutoArc 为false 。在相应的消息处理函数中，对消息处理结果进行确认。
需要注意的是，如果不对消息进行确认，难么MQ会持续的转发消息，最后导致相应的内存暴涨
var consumer = new EventingBasicConsumer(channel);
consumer.Received += (sender, ea) =>
{
var body = ea.Body.ToArray();
var message = Encoding.UTF8.GetString(body);
Console.WriteLine(" [x] Received {0}", message);

int dots = message.Split('.').Length - 1;
Thread.Sleep(dots * 1000);

Console.WriteLine(" [x] Done");

// Note: it is possible to access the channel via
//       ((EventingBasicConsumer)sender).Model here
channel.BasicAck(deliveryTag: ea.DeliveryTag, multiple: false);

};
channel.BasicConsume(queue: “task_queue”, autoAck: false, consumer: consumer);

三．消息的持续

以上我们解决了worker崩溃造成消息的删除。那么 如果Server崩溃造成消息的删除，要根据下面的操作进行

channel.QueueDeclare(queue: “hello”,
durable: true,
exclusive: false,
autoDelete: false,
arguments: null);

var properties = channel.CreateBasicProperties();
properties.Persistent = true;

四．負載均衡的设置
以上的设置 只能保证消息的转发，但是worker 对消息的处理时长可能不同，payload可能不同，为了不将所有的消息平均的分配，而是能够均衡payload

ior we can use the BasicQos method with the prefetchCount = 1 setting.

channel.BasicQos(prefetchSize: 0, prefetchCount: 1, global: false);