Implementation:Kserve Kserve LLMInferenceServiceConfig Minimal CRD

Knowledge Sources	Kserve_Kserve
Domains	Kubernetes, LLM Inference, CRD
Last Updated	2026-02-13 00:00 GMT

Overview

This file defines the minimal CustomResourceDefinition (CRD) for the LLMInferenceServiceConfig custom resource in the serving.kserve.io API group.

Description

Auto-generated by controller-gen (v0.19.0), this CRD provides a minimal-footprint alternative to the full LLMInferenceServiceConfig CRD. The v1alpha1 version uses x-kubernetes-preserve-unknown-fields to skip detailed field validation, while the v1alpha2 version includes partial schema definitions for fields such as baseRefs, model, and parallelism. The resource is namespaced and supports configuration of LLM inference service templates including model specifications, parallelism settings, and pod templates.

Usage

Use this minimal CRD in development or constrained environments where the full CRD (with complete OpenAPI validation) exceeds size limits or where strict field validation is not required. It is applied to the cluster to register the LLMInferenceServiceConfig resource type before creating any LLMInferenceServiceConfig objects.

Code Reference

Source Location

• Repository: Kserve_Kserve
• File: config/crd/minimal/llmisvc/serving.kserve.io_llminferenceserviceconfigs.yaml

Signature

apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
  annotations:
    controller-gen.kubebuilder.io/version: v0.19.0
  name: llminferenceserviceconfigs.serving.kserve.io
spec:
  group: serving.kserve.io
  names:
    kind: LLMInferenceServiceConfig
    listKind: LLMInferenceServiceConfigList
    plural: llminferenceserviceconfigs
    singular: llminferenceserviceconfig
  scope: Namespaced
  versions:
  - name: v1alpha1
    schema:
      openAPIV3Schema:
        properties:
          apiVersion:
            type: string
          kind:
            type: string
          metadata:
            type: object
          spec:
            type: object
            x-kubernetes-map-type: atomic
            x-kubernetes-preserve-unknown-fields: true
        type: object
    served: true
    storage: false
  - name: v1alpha2
    # ... includes partial schema for baseRefs, model, parallelism

Import

kubectl apply -f config/crd/minimal/llmisvc/serving.kserve.io_llminferenceserviceconfigs.yaml

I/O Contract

Field	Type	Description
apiVersion	string	Must be apiextensions.k8s.io/v1
kind	string	Must be CustomResourceDefinition
spec.group	string	API group: serving.kserve.io
spec.names.kind	string	LLMInferenceServiceConfig
spec.scope	string	Namespaced

Versions

Version	Storage	Served	Schema Detail
v1alpha1	false	true	Minimal schema with x-kubernetes-preserve-unknown-fields
v1alpha2	true	true	Partial schema with baseRefs, model, parallelism, template

v1alpha2 Key Spec Fields

Field	Type	Description
spec.baseRefs	array	References to base configuration objects
spec.model	object	Model specification with uri (required), name, and lora adapters
spec.parallelism	object	Parallelism settings: data, dataLocal, dataRPCPort, expert, tensor
spec.template	object	Pod template specification for containers and init containers

Usage Examples

Register the CRD on a Kubernetes cluster:

# Apply the minimal CRD definition
kubectl apply -f config/crd/minimal/llmisvc/serving.kserve.io_llminferenceserviceconfigs.yaml

# Verify the CRD is registered
kubectl get crd llminferenceserviceconfigs.serving.kserve.io

Related Pages

• Kserve_Kserve_LLMInferenceService_Minimal_CRD - Companion minimal CRD for the LLMInferenceService resource
• Kserve_Kserve_LLM_Decode_Worker_DP_Config - Decode worker configuration that creates LLMInferenceServiceConfig objects
• Kserve_Kserve_LLM_Prefill_Worker_DP_Config - Prefill worker configuration that creates LLMInferenceServiceConfig objects
• Kserve_Kserve_LLM_Worker_DP_Config - Standard worker configuration that creates LLMInferenceServiceConfig objects