kafka and kafka-ui, without zookeeper

docker-compose.yaml

name: kafka-sandbox
services:
  kafka:
    image: bitnami/kafka:4.0.0
    environment:
      - KAFKA_CLUSTER_ID=lkorDA4qT6W1K_dk0LHvtg
      # Start Kraft Setup (Kafka as Controller - no Zookeeper)
      - KAFKA_CFG_NODE_ID=1
      - KAFKA_CFG_PROCESS_ROLES=broker,controller
      - KAFKA_CFG_BROKER_ID=1
      - [email protected]:9093
      - KAFKA_CFG_LISTENER_SECURITY_PROTOCOL_MAP=CONTROLLER:PLAINTEXT,PLAINTEXT:PLAINTEXT,INTERNAL:PLAINTEXT
      - KAFKA_CFG_CONTROLLER_LISTENER_NAMES=CONTROLLER
      - KAFKA_CFG_LOG_DIRS=/tmp/logs
      - KAFKA_CFG_LISTENERS=PLAINTEXT://:9092,CONTROLLER://:9093,INTERNAL://:9094
      # End Kraft Specific Setup
      - KAFKA_CFG_ADVERTISED_LISTENERS=PLAINTEXT://127.0.0.1:9092,INTERNAL://kafka:9094
    ports:
      - "0.0.0.0:9092:9092"

  kafka-ui:
    image: provectuslabs/kafka-ui
    ports:
      - "8080:8080"
    restart: "always"
    environment:
      KAFKA_CLUSTERS_0_NAME: "lkorDA4qT6W1K_dk0LHvtg"
      KAFKA_CLUSTERS_0_BOOTSTRAPSERVERS: kafka:9094
    depends_on:
      - kafka

Actually I don't know. I’m using this config only for local development, so I never thought about persistence. However, the Bitnami Docker image page (look at Persisting your data section) suggests a slightly different approach and uses a different path for data inside the container.

kafka:
  ...
  volumes:
    - /path/to/kafka-persistence:/bitnami/kafka
  ...

Thanks again for the help.
I'm using your compose example in production here where I work.

a demonstration of the current data load that kafka handles:

We only have this topic that receives a high volume of messages per hour. This topic is adjusted to retain data for 8 hours.
Only with Kafka can we absorb this data and then insert it into a Postgres DB.
Kafka with 1 worker can support a lot of data, this docker compose is approved 😅