Kafka

Deploy Apache Kafka on Kubernetes using the official apache/kafka:4.2.0 image with a KRaft-only design. The chart intentionally supports a simple single-broker path for development and a production-oriented cluster path with dedicated controllers and brokers.

Key Features

Latest stable Kafka — pinned to the latest stable upstream Kafka release
KRaft only — no ZooKeeper mode in this chart
Two clear topologies — single-broker for simple environments and cluster for production-oriented deployments
Persistent storage — supported for single-broker, controllers, and brokers
Stable broker DNS — advertised listeners use StatefulSet pod DNS names inside the cluster
Optional metrics — JMX exporter javaagent with optional ServiceMonitor

Installation

HTTPS repository:

helm repo add helmforge https://repo.helmforge.dev
helm repo update
helm install kafka helmforge/kafka -f values.yaml

OCI registry:

helm install kafka oci://ghcr.io/helmforgedev/helm/kafka -f values.yaml

Single-Broker Example

architecture: single-broker

singleBroker:
  persistence:
    size: 8Gi

Cluster Example

architecture: cluster

cluster:
  minInSyncReplicas: 2
  controllers:
    replicaCount: 3
    persistence:
      size: 8Gi
  brokers:
    replicaCount: 3
    persistence:
      size: 50Gi

pdb:
  enabled: true

Key Values

Key	Default	Description
`architecture`	`single-broker`	`single-broker` or `cluster`
`image.repository`	`apache/kafka`	Kafka image repository
`image.tag`	`4.2.0`	Kafka image tag
`kraft.existingSecret`	`""`	Existing secret with KRaft cluster ID and controller directory IDs
`listeners.client.port`	`9092`	Kafka client port
`listeners.controller.port`	`9093`	KRaft controller port
`cluster.controllers.replicaCount`	`3`	Dedicated controller replicas
`cluster.brokers.replicaCount`	`3`	Dedicated broker replicas
`cluster.minInSyncReplicas`	`2`	Recommended minimum ISR for cluster mode
`metrics.enabled`	`false`	Enable JMX exporter metrics

Validation Guidance

use single-broker for development and CI, not as the production baseline
use cluster for production-oriented workloads with dedicated controllers and brokers
validate topic creation, producer, and consumer flows before promotion
plan external listeners, TLS, and auth as explicit follow-up work if your environment needs them

More Information

See the source code and full values reference on GitHub.