我正在用django为微服务体系结构构建一个身份验证服务。它更多的是一个实验，而不是一个真实世界的实现，以了解更多的事件源。
我知道django有一个orm，它与框架非常耦合，使用flask将是一个更简单的方法，但我正在尝试找到一种解决方法。
用例非常简单。用户注册后会收到一封电子邮件来激活他的帐户。当一个用户再次被激活时，我发了一封电子邮件通知他他的帐户正在激活。
据我所知，在事件源系统中。事件被触发和存储，我们可以获得最新的状态并将其存储在数据库中。在我的情况下，这将是postgres使用django orm。
我使用django中的信号将事件发布给kafka。pre\u save（）信号。然后模型将保存对象。
如果更新尚未在下面实施。我只会发布更新的字段。并更新django中的对象。
有人看到这种方法有什么警告吗？或者在模型的save方法中实现它会更好吗？
我很想听听你对此的反馈。

# app/services.py

class KafkaService:

    def __init__(self):
        try:
            self.producer = KafkaProducer(bootstrap_servers=settings.KAFKA_BROKERS,
                                          value_serializer=lambda m: json.dumps(m).encode('ascii'))
        except KafkaError as ke:
            logger.exception(f"Something went wrong creating a kafka producer: {ke}")
            self.producer = None
        except Exception as ex:
            logger.exception(f"Something went wrong creating a kafka producer: {ex}")
            self.producer = None

    def publish(self, topic, key, data):
        if not self.producer:
            logger.error(f"Kafka Producer could not establish a connection")
            pass
        try:
            self.producer.send(topic, key=bytes(key, encoding='utf-8'), value=data)
            self.producer.flush()
            logger.info("Message has been published to Kafka")
        except Exception as ex:
            logger.info(f"Errored while publishing to Kafka: {ex}")

# app/events.py

class UserEvent:

    def __init__(self, event_store):  # event store here is Kafka
        """ A user event class which is an injectable. I am using here.
        I need a key for kafka to place the correct event in the correct partition.

        :parameters:
            - event_store: a class form example :class:`KafkaService` which publishes data
        """
        self.event_store = event_store

    def user_created(self, email, data):
        """ Publish an event to the event store when a user is created

        :param email: string
        :param data: string - json
        """
        self.event_store.publish('user-created', email, data)

    def user_activated(self, email, data):
        """ Publish an event to the event store when a user is activated """
        self.event_store.publish('user-activated', email, data)

# app/signals.py

kafka_service = KafkaService()

user_event = UserEvent(kafka_service)

def user_added_event(sender, instance, created,**kwargs):
    if created:
        from users.api.v1.serializers import UserSerializer  # Avoid (Apps Not Read)

        value = UserSerializer(instance).data
        user_event.user_created(instance.email, value)

    else:
        logger.info("User Updated")