Message to a partitioned entity with duplicate detection enabled must have either PartitionKey or MessageId

• Package Name: azure-servicebus
• Package Version: 7.4.0
• Operating System: docker image python:3.8-buster
• Python Version: 3.8.12

Describe the bug I create Standart SB Queue with enabled partitions and duplicate detection.

And for cases, when I send messages in the batch mode, I receive an error Message to a partitioned entity with duplicate detection enabled must have either PartitionKey or MessageId. But when I send the same message one by one - it works ok.

To Reproduce Steps to reproduce the behavior:

1. run this code

connectio_string = 'Endpoint=sb://MY_WORKER_HERE.servicebus.windows.net/;SharedAccessKeyName=ingest_rule;SharedAccessKey=MY_KEY_HERE;EntityPath=MY_QUEUE_NAME_HERE'
uamqp_logger = logging.getLogger('uamqp')
uamqp_logger.setLevel(logging.DEBUG)

_connection_info = parse_connection_string(connectio_string)
_client = ServiceBusClient.from_connection_string(conn_str=connectio_string, logging_enable=True)

with _client.get_queue_sender(_connection_info.entity_path) as sender:
    service_bus_message = sender.create_message_batch()
    service_bus_message.add_message(ServiceBusMessage('msg1', message_id=uuid.uuid4().hex), )
    service_bus_message.add_message(ServiceBusMessage('msg2', message_id=uuid.uuid4().hex), )
    sender.send_messages(service_bus_message)

and the result will be:

ClientMessageError                        Traceback (most recent call last)
/usr/local/lib/python3.8/site-packages/azure/servicebus/_base_handler.py in _do_retryable_operation(self, operation, timeout, **kwargs)
    394                     kwargs["timeout"] = remaining_timeout
--> 395                 return operation(**kwargs)
    396             except StopIteration:

/usr/local/lib/python3.8/site-packages/azure/servicebus/_servicebus_sender.py in _send(self, message, timeout, last_exception)
    261             self._set_msg_timeout(timeout, last_exception)
--> 262             self._handler.send_message(message.message)
    263         finally:  # reset the timeout of the handler back to the default value

/usr/local/lib/python3.8/site-packages/uamqp/client.py in send_message(self, messages, close_on_done)
    743                     details['failed_messages'][failed_message] = exception
--> 744                 raise errors.ClientMessageError(exception, info=details)
    745         finally:

ClientMessageError: ErrorCodes.NotAllowed: Message to a partitioned entity with duplicate detection enabled must have either PartitionKey or MessageId set TrackingId:dbaf8438-ee39-4057-ad37-47371b8749cc_G23, SystemTracker:client-link425450073, Timestamp:2021-12-01T10:19:43

During handling of the above exception, another exception occurred:

ServiceBusError                           Traceback (most recent call last)
<ipython-input-64-489aa3b7946b> in <module>
      6     service_bus_message.add_message(ServiceBusMessage('msg1', message_id=uuid.uuid4()), )
      7     service_bus_message.add_message(ServiceBusMessage('msg2', message_id=uuid.uuid4()), )
----> 8     sender.send_messages(service_bus_message)

/usr/local/lib/python3.8/site-packages/azure/servicebus/_servicebus_sender.py in send_messages(self, message, **kwargs)
    415                 self._add_span_request_attributes(send_span)
    416 
--> 417             self._do_retryable_operation(
    418                 self._send,
    419                 message=obj_message,

/usr/local/lib/python3.8/site-packages/azure/servicebus/_base_handler.py in _do_retryable_operation(self, operation, timeout, **kwargs)
    397                 raise
    398             except Exception as exception:  # pylint: disable=broad-except
--> 399                 last_exception = self._handle_exception(exception)
    400                 if require_last_exception:
    401                     kwargs["last_exception"] = last_exception

/usr/local/lib/python3.8/site-packages/azure/servicebus/_base_handler.py in _handle_exception(self, exception)
    344             self._close_handler()
    345         if not error._retryable:
--> 346             raise error
    347 
    348         return error

ServiceBusError: Message to a partitioned entity with duplicate detection enabled must have either PartitionKey or MessageId set TrackingId:dbaf8438-ee39-4057-ad37-47371b8749cc_G23, SystemTracker:client-link425450073, Timestamp:2021-12-01T10:19:43 Error condition: ErrorCodes.NotAllowed.

Expected behavior Batch sending work as expected for queues, where duplicate detection and partition enabled. Potentially - here is an issue, that batching clear individual broker properties(as message ID) so it should not happen.

[YalinLi0312]

@flying-pi Thanks for your feedback, we’ll investigate asap.

[GinSiuCheng]

@yunhaoling is the above a supported scenario?