[pull] master from kserve:master

.github/actions/minikube-setup/action.yml

@@ -12,7 +12,7 @@ runs:
     - name: Setup Minikube
       uses: medyagh/setup-minikube@latest
       with:
-        minikube-version: '1.31.1'
+        minikube-version: '1.32.0'
         kubernetes-version: 'v1.27.4'
         driver: 'none'
         network-plugin: 'cni'

README.md

@@ -7,20 +7,19 @@
 [![LICENSE](https://img.shields.io/github/license/kserve/kserve.svg)](https://github.com/kserve/kserve/blob/master/LICENSE)
 [![Slack Status](https://img.shields.io/badge/slack-join_chat-white.svg?logo=slack&style=social)](https://github.com/kserve/community/blob/main/README.md#questions-and-issues)
 
-KServe provides a Kubernetes [Custom Resource Definition](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) for serving machine learning (ML) models on arbitrary frameworks. It aims to solve production model serving use cases by providing performant, high abstraction interfaces for common ML frameworks like Tensorflow, XGBoost, ScikitLearn, PyTorch, and ONNX.
+KServe provides a Kubernetes [Custom Resource Definition](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) for serving predictive and generative machine learning (ML) models. It aims to solve production model serving use cases by providing high abstraction interfaces for Tensorflow, XGBoost, ScikitLearn, PyTorch, Huggingface Transformer/LLM models using standardized data plane protocols.
 
 It encapsulates the complexity of autoscaling, networking, health checking, and server configuration to bring cutting edge serving features like GPU Autoscaling, Scale to Zero, and Canary Rollouts to your ML deployments. It enables a simple, pluggable, and complete story for Production ML Serving including prediction, pre-processing, post-processing and explainability. KServe is being [used across various organizations.](https://kserve.github.io/website/master/community/adopters/)
 
 For more details, visit the [KServe website](https://kserve.github.io/website/).
 
-![KServe](/docs/diagrams/kserve.png)
+![KServe](/docs/diagrams/kserve_new.png)
 
-_Since 0.7 [KFServing is rebranded to KServe](https://blog.kubeflow.org/release/official/2021/09/27/kfserving-transition.html), we still support the RTS release
-[0.6.x](https://github.com/kserve/kserve/tree/release-0.6), please refer to corresponding release branch for docs_.
+_Since 0.7 [KFServing is rebranded to KServe](https://blog.kubeflow.org/release/official/2021/09/27/kfserving-transition.html).
 
 ### Why KServe?
-- KServe is a standard, cloud agnostic **Model Inference Platform** on Kubernetes, built for highly scalable use cases.
-- Provides performant, **standardized inference protocol** across ML frameworks.
+- KServe is a standard, cloud agnostic **Model Inference Platform** for serving predictive and generative AI models on Kubernetes, built for highly scalable use cases.
+- Provides performant, **standardized inference protocol** across ML frameworks including OpenAI specification for generative models.
 - Support modern **serverless inference workload** with **request based autoscaling including scale-to-zero** on **CPU and GPU**.
 - Provides **high scalability, density packing and intelligent routing** using **ModelMesh**.
 - **Simple and pluggable production serving** for **inference**, **pre/post processing**, **monitoring** and **explainability**.

charts/kserve-resources/templates/clusterrole.yaml

@@ -63,9 +63,7 @@ rules:
   verbs:
   - create
   - get
-  - list
   - update
-  - watch
 - apiGroups:
   - ""
   resources:
@@ -102,18 +100,14 @@ rules:
   - create
   - delete
   - get
-  - list
   - patch
   - update
-  - watch
 - apiGroups:
   - ""
   resources:
   - serviceaccounts
   verbs:
   - get
-  - list
-  - watch
 - apiGroups:
   - ""
   resources:

cmd/manager/main.go

@@ -23,15 +23,6 @@ import (
 
 	"sigs.k8s.io/controller-runtime/pkg/webhook/admission"
 
-	"github.com/kserve/kserve/pkg/apis/serving/v1alpha1"
-	"github.com/kserve/kserve/pkg/apis/serving/v1beta1"
-	"github.com/kserve/kserve/pkg/constants"
-	graphcontroller "github.com/kserve/kserve/pkg/controller/v1alpha1/inferencegraph"
-	trainedmodelcontroller "github.com/kserve/kserve/pkg/controller/v1alpha1/trainedmodel"
-	"github.com/kserve/kserve/pkg/controller/v1alpha1/trainedmodel/reconcilers/modelconfig"
-	v1beta1controller "github.com/kserve/kserve/pkg/controller/v1beta1/inferenceservice"
-	"github.com/kserve/kserve/pkg/webhook/admission/pod"
-	"github.com/kserve/kserve/pkg/webhook/admission/servingruntime"
 	istio_networking "istio.io/api/networking/v1beta1"
 	istioclientv1beta1 "istio.io/client-go/pkg/apis/networking/v1beta1"
 	v1 "k8s.io/api/core/v1"
@@ -42,13 +33,22 @@ import (
 	"k8s.io/client-go/tools/record"
 	knservingv1 "knative.dev/serving/pkg/apis/serving/v1"
 	ctrl "sigs.k8s.io/controller-runtime"
-	"sigs.k8s.io/controller-runtime/pkg/client"
 	"sigs.k8s.io/controller-runtime/pkg/client/config"
 	"sigs.k8s.io/controller-runtime/pkg/log/zap"
 	"sigs.k8s.io/controller-runtime/pkg/manager"
 	"sigs.k8s.io/controller-runtime/pkg/manager/signals"
 	metricsserver "sigs.k8s.io/controller-runtime/pkg/metrics/server"
 	"sigs.k8s.io/controller-runtime/pkg/webhook"
+
+	"github.com/kserve/kserve/pkg/apis/serving/v1alpha1"
+	"github.com/kserve/kserve/pkg/apis/serving/v1beta1"
+	"github.com/kserve/kserve/pkg/constants"
+	graphcontroller "github.com/kserve/kserve/pkg/controller/v1alpha1/inferencegraph"
+	trainedmodelcontroller "github.com/kserve/kserve/pkg/controller/v1alpha1/trainedmodel"
+	"github.com/kserve/kserve/pkg/controller/v1alpha1/trainedmodel/reconcilers/modelconfig"
+	v1beta1controller "github.com/kserve/kserve/pkg/controller/v1beta1/inferenceservice"
+	"github.com/kserve/kserve/pkg/webhook/admission/pod"
+	"github.com/kserve/kserve/pkg/webhook/admission/servingruntime"
 )
 
 var (
@@ -112,6 +112,13 @@ func main() {
 		os.Exit(1)
 	}
 
+	// Setup clientset to directly talk to the api server
+	clientSet, err := kubernetes.NewForConfig(cfg)
+	if err != nil {
+		setupLog.Error(err, "unable to create clientSet")
+		os.Exit(1)
+	}
+
 	// Create a new Cmd to provide shared dependencies and start components
 	setupLog.Info("Setting up manager")
 	mgr, err := manager.New(cfg, manager.Options{
@@ -142,17 +149,12 @@ func main() {
 		os.Exit(1)
 	}
 
-	kubeClient, err := client.New(mgr.GetConfig(), client.Options{Scheme: mgr.GetScheme()})
-	if err != nil {
-		setupLog.Error(err, "unable to create new client.")
-	}
-
-	deployConfig, err := v1beta1.NewDeployConfig(kubeClient)
+	deployConfig, err := v1beta1.NewDeployConfig(clientSet)
 	if err != nil {
 		setupLog.Error(err, "unable to get deploy config.")
 		os.Exit(1)
 	}
-	ingressConfig, err := v1beta1.NewIngressConfig(kubeClient)
+	ingressConfig, err := v1beta1.NewIngressConfig(clientSet)
 	if err != nil {
 		setupLog.Error(err, "unable to get ingress config.")
 		os.Exit(1)
@@ -181,16 +183,12 @@ func main() {
 	// Setup all Controllers
 	setupLog.Info("Setting up v1beta1 controller")
 	eventBroadcaster := record.NewBroadcaster()
-	clientSet, err := kubernetes.NewForConfig(mgr.GetConfig())
-	if err != nil {
-		setupLog.Error(err, "unable to create clientSet")
-		os.Exit(1)
-	}
 	eventBroadcaster.StartRecordingToSink(&typedcorev1.EventSinkImpl{Interface: clientSet.CoreV1().Events("")})
 	if err = (&v1beta1controller.InferenceServiceReconciler{
-		Client: mgr.GetClient(),
-		Log:    ctrl.Log.WithName("v1beta1Controllers").WithName("InferenceService"),
-		Scheme: mgr.GetScheme(),
+		Client:    mgr.GetClient(),
+		Clientset: clientSet,
+		Log:       ctrl.Log.WithName("v1beta1Controllers").WithName("InferenceService"),
+		Scheme:    mgr.GetScheme(),
 		Recorder: eventBroadcaster.NewRecorder(
 			mgr.GetScheme(), v1.EventSource{Component: "v1beta1Controllers"}),
 	}).SetupWithManager(mgr, deployConfig, ingressConfig); err != nil {
@@ -207,7 +205,7 @@ func main() {
 		Log:                   ctrl.Log.WithName("v1beta1Controllers").WithName("TrainedModel"),
 		Scheme:                mgr.GetScheme(),
 		Recorder:              eventBroadcaster.NewRecorder(mgr.GetScheme(), v1.EventSource{Component: "v1beta1Controllers"}),
-		ModelConfigReconciler: modelconfig.NewModelConfigReconciler(mgr.GetClient(), mgr.GetScheme()),
+		ModelConfigReconciler: modelconfig.NewModelConfigReconciler(mgr.GetClient(), clientSet, mgr.GetScheme()),
 	}).SetupWithManager(mgr); err != nil {
 		setupLog.Error(err, "unable to create controller", "v1beta1Controllers", "TrainedModel")
 		os.Exit(1)
@@ -218,10 +216,11 @@ func main() {
 	setupLog.Info("Setting up InferenceGraph controller")
 	inferenceGraphEventBroadcaster.StartRecordingToSink(&typedcorev1.EventSinkImpl{Interface: clientSet.CoreV1().Events("")})
 	if err = (&graphcontroller.InferenceGraphReconciler{
-		Client:   mgr.GetClient(),
-		Log:      ctrl.Log.WithName("v1alpha1Controllers").WithName("InferenceGraph"),
-		Scheme:   mgr.GetScheme(),
-		Recorder: eventBroadcaster.NewRecorder(mgr.GetScheme(), v1.EventSource{Component: "InferenceGraphController"}),
+		Client:    mgr.GetClient(),
+		Clientset: clientSet,
+		Log:       ctrl.Log.WithName("v1alpha1Controllers").WithName("InferenceGraph"),
+		Scheme:    mgr.GetScheme(),
+		Recorder:  eventBroadcaster.NewRecorder(mgr.GetScheme(), v1.EventSource{Component: "InferenceGraphController"}),
 	}).SetupWithManager(mgr, deployConfig); err != nil {
 		setupLog.Error(err, "unable to create controller", "v1alpha1Controllers", "InferenceGraph")
 		os.Exit(1)
@@ -232,7 +231,7 @@ func main() {
 
 	setupLog.Info("registering webhooks to the webhook server")
 	hookServer.Register("/mutate-pods", &webhook.Admission{
-		Handler: &pod.Mutator{Client: mgr.GetClient(), Decoder: admission.NewDecoder(mgr.GetScheme())},
+		Handler: &pod.Mutator{Client: mgr.GetClient(), Clientset: clientSet, Decoder: admission.NewDecoder(mgr.GetScheme())},
 	})
 
 	setupLog.Info("registering cluster serving runtime validator webhook to the webhook server")

config/default/kustomization.yaml

@@ -27,11 +27,6 @@ replacements:
     name: kserve-webhook-server-service
     version: v1
   targets:
-  - fieldPaths:
-    - spec.conversion.webhook.clientConfig.service.name
-    select:
-      kind: CustomResourceDefinition
-      name: inferenceservices.serving.kserve.io
   - fieldPaths:
     - webhooks.*.clientConfig.service.name
     select:
@@ -78,11 +73,6 @@ replacements:
     name: kserve-controller-manager
     version: v1
   targets:
-  - fieldPaths:
-    - spec.conversion.webhook.clientConfig.service.namespace
-    select:
-      kind: CustomResourceDefinition
-      name: inferenceservices.serving.kserve.io
   - fieldPaths:
     - webhooks.*.clientConfig.service.namespace
     select:
@@ -199,5 +189,4 @@ patches:
 - path: clusterservingruntime_validatingwebhook_cainjection_patch.yaml
 - path: servingruntime_validationwebhook_cainjection_patch.yaml
 - path: manager_resources_patch.yaml
-- path: inferenceservice_conversion_webhook.yaml
 - path: cainjection_conversion_webhook.yaml

config/rbac/role.yaml

@@ -48,9 +48,7 @@ rules:
   verbs:
   - create
   - get
-  - list
   - update
-  - watch
 - apiGroups:
   - ""
   resources:
@@ -87,18 +85,14 @@ rules:
   - create
   - delete
   - get
-  - list
   - patch
   - update
-  - watch
 - apiGroups:
   - ""
   resources:
   - serviceaccounts
   verbs:
   - get
-  - list
-  - watch
 - apiGroups:
   - ""
   resources:

go.mod

@@ -5,13 +5,13 @@ go 1.21
 require (
 	cloud.google.com/go/storage v1.35.1
 	github.com/aws/aws-sdk-go v1.48.0
-	github.com/cloudevents/sdk-go/v2 v2.14.0
+	github.com/cloudevents/sdk-go/v2 v2.15.2
 	github.com/fsnotify/fsnotify v1.7.0
 	github.com/getkin/kin-openapi v0.120.0
 	github.com/go-logr/logr v1.3.0
 	github.com/gofrs/uuid/v5 v5.0.0
 	github.com/google/go-cmp v0.6.0
-	github.com/google/uuid v1.4.0
+	github.com/google/uuid v1.6.0
 	github.com/googleapis/google-cloud-go-testing v0.0.0-20210719221736-1c9a4c676720
 	github.com/json-iterator/go v1.1.12
 	github.com/kelseyhightower/envconfig v1.4.0
@@ -22,7 +22,7 @@ require (
 	github.com/spf13/pflag v1.0.5
 	github.com/stretchr/testify v1.8.4
 	github.com/tidwall/gjson v1.17.0
-	go.uber.org/zap v1.26.0
+	go.uber.org/zap v1.27.0
 	gomodules.xyz/jsonpatch/v2 v2.4.0
 	google.golang.org/api v0.151.0
 	google.golang.org/protobuf v1.32.0

go.sum

@@ -76,6 +76,8 @@ github.com/chzyer/test v0.0.0-20180213035817-a1ea475d72b1/go.mod h1:Q3SI9o4m/ZMn
 github.com/client9/misspell v0.3.4/go.mod h1:qj6jICC3Q7zFZvVWo7KLAzC3yx5G7kyvSDkc90ppPyw=
 github.com/cloudevents/sdk-go/v2 v2.14.0 h1:Nrob4FwVgi5L4tV9lhjzZcjYqFVyJzsA56CwPaPfv6s=
 github.com/cloudevents/sdk-go/v2 v2.14.0/go.mod h1:xDmKfzNjM8gBvjaF8ijFjM1VYOVUEeUfapHMUX1T5To=
+github.com/cloudevents/sdk-go/v2 v2.15.2 h1:54+I5xQEnI73RBhWHxbI1XJcqOFOVJN85vb41+8mHUc=
+github.com/cloudevents/sdk-go/v2 v2.15.2/go.mod h1:lL7kSWAE/V8VI4Wh0jbL2v/jvqsm6tjmaQBSvxcv4uE=
 github.com/cncf/udpa/go v0.0.0-20191209042840-269d4d468f6f/go.mod h1:M8M6+tZqaGXZJjfX53e64911xZQV5JYwmTeXPW+k8Sc=
 github.com/cpuguy83/go-md2man/v2 v2.0.3/go.mod h1:tgQtvFlXSQOSOSIRvRPT7W67SCa46tRHOmNcaadrF8o=
 github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ33E=
@@ -211,6 +213,8 @@ github.com/google/s2a-go v0.1.7/go.mod h1:50CgR4k1jNlWBu4UfS4AcfhVe1r6pdZPygJ3R8
 github.com/google/uuid v1.1.2/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/google/uuid v1.4.0 h1:MtMxsa51/r9yyhkyLsVeVt0B+BGQZzpQiTQ4eHZ8bc4=
 github.com/google/uuid v1.4.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
+github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
+github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/googleapis/enterprise-certificate-proxy v0.3.2 h1:Vie5ybvEvT75RniqhfFxPRy3Bf7vr3h0cechB90XaQs=
 github.com/googleapis/enterprise-certificate-proxy v0.3.2/go.mod h1:VLSiSSBs/ksPL8kq3OBOQ6WRI2QnaFynd1DCjZ62+V0=
 github.com/googleapis/gax-go/v2 v2.0.4/go.mod h1:0Wqv26UfaUD9n4G6kQubkQ+KchISgw+vpHVxEJEs9eg=
@@ -397,6 +401,8 @@ go.uber.org/multierr v1.11.0/go.mod h1:20+QtiLqy0Nd6FdQB9TLXag12DsQkrbs3htMFfDN8
 go.uber.org/zap v1.10.0/go.mod h1:vwi/ZaCAaUcBkycHslxD9B2zi4UTXhF60s6SWpuDF0Q=
 go.uber.org/zap v1.26.0 h1:sI7k6L95XOKS281NhVKOFCUNIvv9e0w4BF8N3u+tCRo=
 go.uber.org/zap v1.26.0/go.mod h1:dtElttAiwGvoJ/vj4IwHBS/gXsEu/pZ50mUIRWuG0so=
+go.uber.org/zap v1.27.0 h1:aJMhYGrd5QSmlpLMr2MftRKl7t8J8PTZPA732ud/XR8=
+go.uber.org/zap v1.27.0/go.mod h1:GB2qFLM7cTU87MWRP2mPIjqfIDnGu+VIO4V/SdhGo2E=
 golang.org/x/crypto v0.17.0 h1:r8bRNjWL3GshPW3gkd+RpvzWrZAwPS49OmTGZ/uhM4k=
 golang.org/x/crypto v0.17.0/go.mod h1:gCAAfMLgwOJRpTjQ2zCCt2OcSfYMTeZVSRtQlPC7Nq4=
 golang.org/x/exp v0.0.0-20190121172915-509febef88a4/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=

hack/quick_install.sh

@@ -30,10 +30,10 @@ while getopts ":hsr" option; do
    esac
 done
 
-export ISTIO_VERSION=1.19.4
+export ISTIO_VERSION=1.20.4
 export ISTIO_DIR=istio-${ISTIO_VERSION}
-export KNATIVE_SERVING_VERSION=knative-v1.10.1
-export KNATIVE_ISTIO_VERSION=knative-v1.10.0
+export KNATIVE_SERVING_VERSION=knative-v1.13.1
+export KNATIVE_ISTIO_VERSION=knative-v1.13.1
 export KSERVE_VERSION=v0.12.0
 export CERT_MANAGER_VERSION=v1.9.0
 export SCRIPT_DIR="$( dirname -- "${BASH_SOURCE[0]}" )"