我正在尝试使用rabbitmq-delayed-message-exchange插件,该插件似乎可以正常发送延迟消息,该消息将在延迟间隔后收到。在消费消息时,我应该在rabbitmq tutorial 2中实现通常的循环方法,但我收到的是多个消费者消费相同的消息。客户端功能是我在做basic.consume的地方。它与循环教程中的代码和延迟插件大致相同。
代码的工作方式如下所示 - 接收函数接收一条消息,该消息具有特定的延迟,将在队列中发回以供以后使用。 发送消息以延迟发送消息。
我面临的问题是使用basic.consume,由于一些配置错误,我想我会让多个消费者使用相同的消息。
我是否需要更改以实现多个消费者避免处理相同请求。
class cities:
def __init__(self):
self.connection = pika.BlockingConnection(pika.ConnectionParameters('localhost'))
self.channel = self.connection.channel()
self.channel.exchange_declare("test-x", type="x-delayed-message", arguments={"x-delayed-type":"direct"})
self.channel.queue_declare(queue='towns_queue')
self.channel.queue_bind(queue="towns_queue", exchange="test-x", routing_key="towns_queue")
self.logger = logging.getLogger("rabbitmq")
self.logger.setLevel(logging.DEBUG)
handler = RotatingFileHandler("logs/Receive.log", maxBytes=100*1024*1024, backupCount=100)
self.logger.addHandler(handler)
formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
handler.setFormatter(formatter)
self.errlogger = logging.getLogger("Msg_timer_Error")
self.errlogger.setLevel(logging.ERROR)
errhandler = RotatingFileHandler("logs/towns_error.log", maxBytes=100*1024*1024, backupCount=100)
self.errlogger.addHandler(errhandler)
errformatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
errhandler.setFormatter(errformatter)
def call_api(self,url):
do something
def server_log(self,msg,value):
self.logger.info(msg+" : "+str(value))
def send_msg(self,msg,delay):
try:
delay = delay - 40
self.server_log("Delay Set : ",delay)
delay *=1000
self.channel.basic_publish(exchange='test-x',routing_key='towns_queue',body=msg,properties=pika.BasicProperties(headers={"x-delay"\
:delay},delivery_mode=1))
self.server_log("Msg Sent",msg)
except:
self.errlogger.exception(msg)
self.errlogger.exception(str(traceback.print_exc()))
pass
def receive(self,ch, method, properties, body):
try:
self.server_log("Received message : ",body)
self.process(body)
time.sleep(1)
ch.basic_ack(delivery_tag = method.delivery_tag)
return body
except:
self.errlogger.exception(body)
self.errlogger.exception(str(traceback.print_exc()))
pass
def process(self,body):
delay=self.call_api(body)
if delay:
self.send_msg(body,delay)
def client(self):
try:
self.channel.basic_qos(prefetch_count=1)
self.channel.basic_consume(self.receive,queue='towns_queue',no_ack=False)
self.channel.start_consuming()
except:
self.errlogger.exception(str(traceback.print_exc()))
pass
def close(self):
try:
self.connection.close()
except:
self.errlogger.exception(str(traceback.print_exc()))
pass
def multiprocess(num_process=4):
procs = []
objs = []
for i in range(4):
p = multiprocessing.Process(target=cities().client)
p.start()
procs.append(p)
for p in procs:
p.join()
if __name__ == "__main__":
if len(sys.argv)> 1:
num_proc = sys.argv[1]
if num_proc.isdigit():
num_proc = int(num_proc)
multiprocess(num_proc)
else:
raise Exception ('num_of_processes needs to be an integer.')
else:
multiprocess()
感谢任何帮助。
更新:
这可能有问题:
def client(self):
try:
self.channel.basic_qos(prefetch_count=1)
self.channel.basic_consume(self.receive,queue='towns_queue',no_ack=False)
self.channel.start_consuming()
except:
self.errlogger.exception(str(traceback.print_exc()))
pass
答案 0 :(得分:0)
本教程中的示例为每条消息创建一个新连接。我尝试通过改变我的send_message函数来这样工作,在那里我将创建一个新的客户端对象,它将创建一个新的连接来发送并在完成后关闭连接。所以,一旦改变了,它就会起作用。
ERROR: (gcloud.compute.forwarding-rules.describe) Could not fetch resource:
- The resource 'projects/my-project/regions/us-central1/forwardingRules/my-forwarding' was not found
此外,必须更改__init__构造函数以避免每次都创建记录器,但这不是问题的一部分。