Implementation:Kserve Kserve ServingRuntime CRD

Knowledge Sources	Kserve_Kserve KServe Docs
Domains	Kubernetes, CRD, Serving Runtime
Last Updated	2026-02-13 00:00 GMT

Overview

Concrete CRD definition for the ServingRuntime custom resource in the KServe serving API.

Description

This file contains the auto-generated full CustomResourceDefinition for the ServingRuntime kind, produced by controller-gen v0.19.0. It belongs to the serving.kserve.io API group at version v1alpha1 and is a namespaced resource. The CRD is structurally identical to ClusterServingRuntime but scoped to a single namespace, allowing namespace-level serving runtime templates that can override or supplement cluster-wide definitions. It includes printer columns for Disabled status, ModelType, Containers, and Age.

Usage

Apply this CRD during KServe installation to register the ServingRuntime API with the Kubernetes API server. Namespace administrators then create ServingRuntime resources to define namespace-scoped serving runtime templates that InferenceService resources within the same namespace can reference. Namespace-scoped runtimes take precedence over cluster-scoped ClusterServingRuntime resources of the same name.

Code Reference

Source Location

• Repository: Kserve_Kserve
• File: config/crd/full/serving.kserve.io_servingruntimes.yaml
• Lines: 1-4191

Signature

apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
  annotations:
    controller-gen.kubebuilder.io/version: v0.19.0
  name: servingruntimes.serving.kserve.io
spec:
  group: serving.kserve.io
  names:
    kind: ServingRuntime
    listKind: ServingRuntimeList
    plural: servingruntimes
    singular: servingruntime
  scope: Namespaced
  versions:
    - name: v1alpha1
      additionalPrinterColumns:
        - jsonPath: .spec.disabled
          name: Disabled
          type: boolean
        - jsonPath: .spec.supportedModelFormats[*].name
          name: ModelType
          type: string
        - jsonPath: .spec.containers[*].name
          name: Containers
          type: string
        - jsonPath: .metadata.creationTimestamp
          name: Age
          type: date
      subresources: {}

Import

kubectl apply -f config/crd/full/serving.kserve.io_servingruntimes.yaml

I/O Contract

Inputs

Name	Type	Required	Description
apiVersion	string	Yes	Must be serving.kserve.io/v1alpha1
kind	string	Yes	Must be ServingRuntime
metadata	ObjectMeta	Yes	Standard Kubernetes object metadata including namespace
spec	ServingRuntimeSpec	Yes	Serving runtime specification defining containers, model formats, and runtime behavior

Key spec fields:

Field	Type	Required	Description
spec.containers	[]Container	Yes	One or more container definitions for the serving runtime (image, command, args, env, resources, ports, volume mounts)
spec.supportedModelFormats	[]SupportedModelFormat	No	List of model formats supported by this runtime (e.g., sklearn, tensorflow, pytorch, onnx)
spec.disabled	boolean	No	When true, this runtime is disabled and will not be selected for model serving
spec.affinity	Affinity	No	Pod scheduling affinity rules (node affinity, pod affinity, pod anti-affinity)
spec.tolerations	[]Toleration	No	Pod tolerations for scheduling on tainted nodes
spec.volumes	[]Volume	No	Additional volumes to mount into the runtime containers

Outputs

Name	Type	Description
status	object	Empty status object (no status fields are currently defined for this resource)

Usage Examples

Create a ServingRuntime

apiVersion: serving.kserve.io/v1alpha1
kind: ServingRuntime
metadata:
  name: custom-sklearn-runtime
  namespace: ml-team
spec:
  supportedModelFormats:
    - name: sklearn
      version: "1"
      autoSelect: true
  containers:
    - name: kserve-container
      image: myregistry/sklearn-server:v1.2
      args:
        - --model_name={{.Name}}
        - --model_dir=/mnt/models
      resources:
        requests:
          cpu: 500m
          memory: 512Mi
        limits:
          cpu: "1"
          memory: 1Gi

List ServingRuntimes

kubectl get servingruntimes -n ml-team
# NAME                     DISABLED   MODELTYPE   CONTAINERS         AGE
# custom-sklearn-runtime   false      sklearn     kserve-container   3d

Related Pages

• Kserve_Kserve_ClusterServingRuntime_CRD -- Cluster-scoped counterpart of this namespace-scoped resource
• New principle needed: Kserve_Kserve_ServingRuntime_Specification