A process is launched within the callback function, and an error occurs when attempting to send an MQ message within that process.

i write a code like that:
```python
from multiprocessing import Process
from rocketmq.client import Producer, Message, PushConsumer, ConsumeStatus
import time
import json

def rmq_consumer():
consumer = PushConsumer('group_demo')
consumer.set_name_server_address('10.13.2.12:9876')
consumer.subscribe('start_or_end_detect', send_message, "*")
print(' [Consumer] Waiting for messages.')
consumer.start()
while True:
time.sleep(1)

def start_pro():
producer = Producer('test_producer')
producer.set_name_server_address('10.13.2.12:9876')
producer.start()
try:
event = {"aa": "vv", "bb": 33}
msg = Message("fireMsg")
ss = json.dumps(event).encode('utf-8')
msg.set_body(ss)
producer.send_sync(msg)

finally:
producer.shutdown()

def send_message():
i = Process(target=start_pro)
i.start()

return ConsumeStatus.CONSUME_SUCCESS

if __name__ == '__main__':
rmq_con = Process(target=rmq_consumer)
```
The code snippet "'**producer.send_sync(msg)**'"
throws an error
**"rocketmq.exceptions.ProducerSendSyncFailed: No route info of this topic: fireMsg,error:-1,in file rocketmq-client-cpp/src/producer/DefaultMQProducerImpl.cpp line:434"**

What could be the cause of this? My RocketMQ deployment is functioning normally.

**If I don't execute Process(target=start_pro) within the send_message method and instead directly include the contents of start_pro within send_message, the aforementioned error does not occur.**

Looking forward to your response. Thank you.

首先重現圍繞 rmq_consumer、send_message、start_pro 和 producer.send_sync 的範例，並比較直接 callback 路徑與 Process(target=start_pro) 路徑。追蹤 Python 用戶端及其 RocketMQ 用戶端相依套件中的 producer 啟動與 topic 路由查找過程。完成標準是確定子程序為何沒有收到 fireMsg 的路由資訊，並記錄經過驗證的原因或修正方案。

由索引模型根據 Issue 內容生成。