kafka.errors.NodeNotReadyError: NodeNotReadyError

[StudyForZX]

kafka-python=2.0.1

kafka_version=2.5.0

code is

from kafka.admin import KafkaAdminClient

self.adminClient = KafkaAdminClient(
            bootstrap_servers = 'borker:9092',
            api_version = (0, 10),
            security_protocol = 'PLAINTEXT',
        )

kafka and zk has no problems.

[mmenbawy]

Check this https://github.com/dpkp/kafka-python/issues/1861 and this https://www.confluent.io/blog/kafka-listeners-explained/

tl;dr: You need to set advertised.listeners (or KAFKA_ADVERTISED_LISTENERS if you’re using Docker images) to the external address (host/IP) so that clients can correctly connect to it. Otherwise, they’ll try to connect to the internal host address—and if that’s not reachable, then problems ensue.

[AntmanTmp]

kafka-python version

crc32c==2.2
kafka-python==2.0.2
kazoo==2.8.0
  - six [required: Any, installed: 1.15.0]

partial docker-compose file

 environment:
      KAFKA_BROKER_ID: 2
      KAFKA_ADVERTISED_HOST_NAME: 10.0.75.2                 
      KAFKA_ADVERTISED_PORT: 9094                              
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://10.0.75.2:9094  
      KAFKA_ZOOKEEPER_CONNECT: "zookeeper1:2181,zookeeper2:2181,zookeeper3:2181"

This is my code below

class KafkaDemo(object):

    @classmethod
    def admin(cls):
        client = KafkaAdminClient(
            bootstrap_servers=[
                '127.0.0.1:9093',
                '127.0.0.1:9094',
                '127.0.0.1:9095'
            ],
            client_id='client-two'
        )
        print(f'Client config: {client.config}')
    
        cluster=client.describe_cluster()

        #topics

        t1 = NewTopic(name='topic1-test', num_partitions=3,replication_factor=1)
        client.create_topics([t1])
        t_info=client.describe_topics(['topic1-test'])
        print(t_info)
        client.delete_topics(['topic1-test'])
        
		client.close()
		
if __name__ == '__main__':
    KafkaDemo.admin()

Error

Connected to pydev debugger (build 202.7660.27)
Traceback (most recent call last):
  File "C:\Users\Administrator\.virtualenvs\distrubution-bUBhtHd3\lib\site-packages\kafka\client_async.py", line 909, in check_version
    version = conn.check_version(timeout=remaining, strict=strict, topics=list(self.config['bootstrap_topics_filter']))
  File "C:\Users\Administrator\.virtualenvs\distrubution-bUBhtHd3\lib\site-packages\kafka\conn.py", line 1240, in check_version
    raise Errors.NodeNotReadyError()
kafka.errors.NodeNotReadyError: NodeNotReadyError
python-BaseException

After the first connection is successful, it can run normally, but once the program reports an error midway, after closing the program, the connection will report an error again

[cuiyuanyuan542]

My environment:

• Kubernetes v1.23.7
• Deploy kafka&zookeeper with YAML I set the NodePort to the kafka Service. But I can't use NodeIP:NodePort as bootstrap-server address. It report same error: kafka.errors.NoBrokersAvailable: NoBrokersAvailable

[cuiyuanyuan542]

My environment:

• Kubernetes v1.23.7
• Deploy kafka&zookeeper with YAML I set the NodePort to the kafka Service. But I can't use NodeIP:NodePort as bootstrap-server address. It report same error: kafka.errors.NoBrokersAvailable: NoBrokersAvailable

I don't know how the kafka env variables in YAML add