设置Pika ioloop异步超时(RabbitMQ)

Question

我需要能够优雅地阻止在Pika ioloop中工作的消费者(工人)。工作人员应在60秒后停止。当前处理的消息应该已完成。

我试图在回调函数中放入一个connection.close()，但这只会停止当前线程，而不会停止整个ioloop。它给出了一个可怕的错误输出。

请参见我的代码中的第16行和以下内容:我使用了(关于Pika ioloop http://pika.github.com/connecting.html#cps-example的基本示例

    from pika.adapters import SelectConnection
    channel = None
    def on_connected(connection):
        connection.channel(on_channel_open)

    def on_channel_open(new_channel):
        global channel
        channel = new_channel
        channel.queue_declare(queue="test", durable=True, exclusive=False, auto_delete=False, callback=on_queue_declared)

    def on_queue_declared(frame):
        channel.basic_consume(handle_delivery, queue='test')

    def handle_delivery(channel, method, header, body):
        print body

        # timer stuff which did NOT work
        global start_time, timeout, connection
        time_diff = time.time()-start_time
        if time_diff > timeout:
            #raise KeyboardInterrupt
            connection.close()

    timeout = 60
    start_time = time.time()

    connection = SelectConnection(parameters, on_connected)

    try:
        connection.ioloop.start()
    except KeyboardInterrupt:
        connection.close()
        connection.ioloop.start()

Answer 1

您可以在打开的连接上附加超时回调函数。下面是您的示例的额外代码。

timeout = 60

def on_timeout():
  global connection
  connection.close()

connection.add_timeout(timeout, on_timeout)

Answer 2

您可以尝试使用：

connection.ioloop.stop()