Kafka does not have a web interface. Interaction is directly between producers and consumers using the broker's port.

‍

Connection string

<IP_DE_LA_VM>:30909

‍

Verify connectivity with kafka-topics

# List existing topics
kafka-topics.sh --bootstrap-server <VM_IP>:30909 --list

# Create a topic
kafka-topics.sh --bootstrap-server <VM_IP>:30909 \ 
--create --topic my-topic --partitions 3 --replication-factor 1

# Describe a topic
kafka-topics.sh --bootstrap-server <VM_IP>:30909 \ 
--describe --topic my-topic

‍

Example with producer and consumer

# Send messages (producer)
kafka-console-producer.sh --bootstrap-server <VM_IP>:30909 --topic my-topic

# Receive messages (consumer)
kafka-console-consumer.sh --bootstrap-server <VM_IP>:30909

-topic my-topic --from-beginning

‍

Connection string for applications

# Java / Spring Boot
bootstrap-servers=<IP_DE_LA_VM>:30909

# Python (kafka-python)
bootstrap_servers=['<IP_DE_LA_VM>:30909']

# Node.js (kafkajs)
brokers: ['<IP_DE_LA_VM>:30909']

# View the Kafka pod
`kubectl get pods -A | grep kafka`

# View the installation log
`tail -f /var/log/cuemby/bootstrap.log`

# Verify that the port is listening
`ss -tlnp | grep 30909`

‍

Expected output:

NAME                     READY   STATUS    RESTARTS
kafka-xxxxxxxxx-xxxxx    1/1     Running   0        ← Running ✓

‍

ℹ️ This installation uses KRaft mode (without ZooKeeper). It does not require a separate ZooKeeper ensemble.

# 1. Create a topic with 3 partitions
kafka-topics.sh --bootstrap-server <VM_IP>:30909
--create --topic events --partitions 3 --replication-factor 1

# 2. Verify that the topic was created
kafka-topics.sh --bootstrap-server <VM_IP>:30909 --list

# 3. Produce a test message
echo "hello world" | kafka-console-producer.sh
--bootstrap-server <VM_IP>:30909 --topic events

# 4. Consume the message
kafka-console-consumer.sh --bootstrap-server <VM_IP>:30909
--topic events --from-beginning --max-messages 1

‍

💡 Tip: Adjust KAFKA_HEAP_OPTS if you see OutOfMemoryError errors. With 8 GB of RAM, a value like -Xmx3g or -Xms3g provides enough headroom for light production workloads.