New to KubeVault? Please start here.

KafkaRole

What is KafkaRole

A KafkaRole is a Kubernetes CustomResourceDefinition (CRD) which allows a user to create a database secret engine role in a Kubernetes native way.

When a KafkaRole is created, the KubeVault operator creates a role according to specification. If the user deletes the KafkaRole CRD, then the respective role will also be deleted from Vault.

KafkaRole CRD Specification

Like any official Kubernetes resource, a KafkaRole object has TypeMeta, ObjectMeta, Spec and Status sections.

A sample KafkaRole object is shown below:

apiVersion: engine.kubevault.com/v1alpha1
kind: KafkaRole
metadata:
  name: kafka-role
  namespace: demo
spec:
  secretEngineRef:
    name: vault-app
  creationStatements:
    - '{"mechanism":"SCRAM-SHA-256","acls":[{"resource_type":"TOPIC","resource_name":"*","operation":"ALL","permission":"ALLOW"},{"resource_type":"GROUP","resource_name":"*","operation":"ALL","permission":"ALLOW"}]}'
status:
  observedGeneration: 1
  phase: Success

Note: To resolve the naming conflict, name of the role in Vault will follow this format: k8s.{clusterName}.{metadata.namespace}.{metadata.name}

Here, we are going to describe the various sections of the KafkaRole crd.

KafkaRole Spec

KafkaRole spec contains information that necessary for creating a database role.

spec:
  secretEngineRef:
    name: <vault-appbinding-name>
  defaultTTL: <default-ttl>
  maxTTL: <max-ttl>
  creationStatements:
    - "statement-0"

KafkaRole spec has the following fields:

spec.secretEngineRef

spec.secretEngineRef is a required field that specifies the name of a SecretEngine.

spec:
  secretEngineRef:
    name: kafka-secret-engine

spec.creationStatements

spec.creationStatements is a required field that specifies a JSON role document used to configure the SCRAM authentication mechanism and optional ACL permissions. The exact schema depends on the plugin — see the example below.

spec:
  creationStatements:
    - '{"mechanism":"SCRAM-SHA-256","acls":[{"resource_type":"TOPIC","resource_name":"*","operation":"ALL","permission":"ALLOW"},{"resource_type":"GROUP","resource_name":"*","operation":"ALL","permission":"ALLOW"}]}'

spec.defaultTTL

spec.defaultTTL is an optional field that specifies the TTL for the leases associated with this role. Accepts time suffixed strings (“1h”) or an integer number of seconds. Defaults to system/engine default TTL time.

spec:
  defaultTTL: "1h"

spec.maxTTL

spec.maxTTL is an optional field that specifies the maximum TTL for the leases associated with this role. Accepts time suffixed strings (“1h”) or an integer number of seconds. Defaults to system/engine default TTL time.

spec:
  maxTTL: "1h"

KafkaRole Status

status shows the status of the KafkaRole. It is managed by the KubeVault operator. It contains the following fields:

• observedGeneration: Specifies the most recent generation observed for this resource. It corresponds to the resource’s generation, which is updated on mutation by the API Server.

• phase: Indicates whether the role successfully applied to Vault or not.

• conditions : Represent observations of a KafkaRole.

Namespace inheritance (tenant isolation)

A KafkaRole never sets or resolves an OpenBao namespace itself. It always inherits the effective namespace of the SecretEngine it references via spec.secretEngineRef (SecretEngine.status.effectiveNamespace) — empty for root, or the tenant’s OpenBao namespace once tenant isolation has placed that engine in one. Every credential this role issues, and every revocation (SecretAccessRequest), is scoped to that same namespace automatically — no KafkaRole-level configuration is needed.