refactor kfserving and upgrade api to latest

.gitignore

@@ -9,4 +9,5 @@ Library
 sdk/arena-python-sdk/dist/
 sdk/arena-python-sdk/build/
 sdk/arena-python-sdk/arenasdk.egg-info/
-__pycache__
+__pycache__
+vendor

charts/kfserving/templates/inferenceservice.yaml

@@ -1,54 +1,9 @@
-{{- $gpuCount := .Values.gpuCount -}}
-apiVersion: serving.kubeflow.org/v1alpha2
-kind: InferenceService
+apiVersion: "serving.kserve.io/v1beta1"
+kind: "InferenceService"
 metadata:
-  annotations:
-  {{- range $key, $value := .Values.annotations }}
-    {{ $key }}: {{ $value | quote }}
-  {{- end }}
-  labels:
-    controller-tools.k8s.io: "1.0"
-    serviceName: "{{ .Values.servingName }}"
-    servingType: "kf-serving"
-    servingName: "{{ .Values.servingName }}"
-    servingVersion: "{{ .Values.servingVersion }}"
-  {{- range $key, $value := .Values.labels }}
-    {{ $key }}: {{ $value | quote }}
-  {{- end }}  
   name: {{ .Release.Name }}
 spec:
-  default:
-    predictor:
-      {{ .Values.modelType }}:
-        storageUri: {{ .Values.storageUri }}
-        container:
-          name: {{ .Release.Name }}
-          image: "{{ .Values.image }}"
-          imagePullPolicy: "{{ .Values.imagePullPolicy }}"
-          {{- if .Values.command }}
-          command: { .Values.command }}
-          {{- end }}
-          ports:
-            - containerPort: {{ .Values.port }}
-              protocol: "TCP"
-          resources:
-            limits:
-              {{- if .Values.cpu }}
-              cpu: {{ .Values.cpu }}
-              {{- end }}
-              {{- if .Values.memory }}
-              memory: {{ .Values.memory }}
-              {{- end }}
-              {{- if gt (int $gpuCount) 0}}
-              nvidia.com/gpu: {{ .Values.gpuCount }}
-              {{- end }}
-            requests:
-              {{- if .Values.cpu }}
-              cpu: {{ .Values.cpu }}
-              {{- end }}
-              {{- if .Values.memory }}
-              memory: {{ .Values.memory }}
-              {{- end }}
-              {{- if gt (int $gpuCount) 0}}
-              nvidia.com/gpu: {{ .Values.gpuCount }}
-              {{- end }}
+  predictor:
+    minReplicas: {{ .Values.minReplicas }}
+    {{ .Values.modelType }}:
+      storageUri: {{ .Values.storageUri }}

charts/kfserving/values.yaml

@@ -10,10 +10,11 @@ image: codait/max-object-detector
 cpu: 100m
 memory: 1Gi
 gpuCount: 0
-storageUri: "hdfs://kfserving-samples/models/tensorflow/flowers-2"
+storageUri: "gs://kfserving-samples/models/sklearn/iris"
 canaryTrafficPercent: 0
 command:
 modelType: custom
+minReplicas: 0
 
 imagePullSecrets: []
 

docs/serving/index.md

@@ -28,7 +28,7 @@ If you want to use arena to manage serving jobs, this guide is for you. we have
 
 ## KFServing Job Guide
 
-* I want to [submit a kfserving job whose type is custom](kfserving/custom.md).
+* I want to [submit a kfserving job](kfserving/kserve.md).
 
 ## Seldon Core Serving Job Guide
 

docs/serving/kfserving/kserve.md

@@ -0,0 +1,61 @@
+# KFServing job 
+
+This guide walks through the steps to deploy and serve a  model with kfserving
+
+1\. Setup
+
+Follow the [KFserving Guide](https://github.com/kubeflow/kfserving#install-kfserving) to install kFserving.For the prerequisites,you should ensure 8g memery and 4 core cpu avaliable in your environment.
+
+2\. summit your serving job into kfserving
+
+    $ arena serve kfserving \
+      >         --name=sklearn-demo \
+      >         --storage-uri="https://github.com/tduffy000/kfserving-uri-examples/blob/master/sklearn/frozen/model.joblib?raw=true" \
+      >         --model-type=sklearn
+      INFO[0000] The Job sklearn-demo has been submitted successfully 
+      inferenceservice.serving.kserve.io/sklearn-demo-202110301430 created
+      INFO[0009] The Job sklearn-demo has been submitted successfully 
+
+
+
+3\. list the job you just serving
+
+    $ arena serve ls                                                                        
+     NAME                       TYPE       VERSION  DESIRED  AVAILABLE  ADDRESS  PORTS
+     sklearn-demo-202110301430  KFServing  1428228  0        0                   N/A
+
+
+4\. test the model service
+
+step1: Determine the ingress IP and ports
+
+The first step is to [determine the ingress IP](https://github.com/kubeflow/kfserving/blob/master/README.md#determine-the-ingress-ip-and-ports) and ports and set INGRESS_HOST and INGRESS_PORT.
+
+This example uses the [codait/max-object-detector](https://github.com/IBM/MAX-Object-Detector) image. The Max Object Detector api server expects a POST request to the /model/predict endpoint that includes an image multipart/form-data and an optional threshold query string.
+
+    MODEL_NAME=sklearn-demo-202110301430
+    SERVICE_HOSTNAME=$(kubectl get inferenceservice ${MODEL_NAME} -o jsonpath='{.status.url}' | cut -d "/" -f 3)
+    INGRESS_HOST=localhost
+    INGRESS_PORT=80
+    curl -v -H "Host: ${SERVICE_HOSTNAME}" http://${INGRESS_HOST}:${INGRESS_PORT}/v1/models/$MODEL_NAME:predict -d '{"instances": [[6.8,  2.8,  4.8,  1.4],[6.0,  3.4,  4.5,  1.6]]}' 
+
+    *   Trying ::1...
+    * TCP_NODELAY set
+    * Connected to localhost (::1) port 80 (#0)
+    > POST /v1/models/sklearn-demo-202110301430:predict HTTP/1.1
+    > Host: sklearn-demo-202110301430.default.example.com
+    > User-Agent: curl/7.64.1
+    > Accept: */*
+    > Content-Length: 64
+    > Content-Type: application/x-www-form-urlencoded
+    > 
+    * upload completely sent off: 64 out of 64 bytes
+    < HTTP/1.1 200 OK
+    < content-length: 23
+    < content-type: application/json; charset=UTF-8
+    < date: Sat, 30 Oct 2021 06:49:17 GMT
+    < server: istio-envoy
+    < x-envoy-upstream-service-time: 10
+    < 
+    * Connection #0 to host localhost left intact
+    {"predictions": [1, 1]}* Closing connection 0

go.mod

@@ -3,33 +3,39 @@ module github.com/kubeflow/arena
 go 1.12
 
 require (
-	github.com/docker/docker v0.7.3-0.20190327010347-be7ac8be2ae0
-	github.com/go-openapi/spec v0.19.3
+	github.com/docker/docker v20.10.7+incompatible
+	github.com/go-openapi/spec v0.20.2
 	github.com/golang/glog v0.0.0-20160126235308-23def4e6c14b
-	github.com/golang/protobuf v1.4.2
+	github.com/golang/protobuf v1.5.2
 	github.com/google/uuid v1.3.0
 	github.com/mitchellh/go-homedir v1.1.0
-	github.com/prometheus/client_golang v1.0.0
-	github.com/prometheus/common v0.4.1
-	github.com/sirupsen/logrus v1.4.2
-	github.com/spf13/cobra v0.0.5
-	github.com/spf13/viper v1.7.1
-	github.com/stretchr/testify v1.4.0
+	github.com/moby/term v0.0.0-20210619224110-3f7ff695adc6 // indirect
+	github.com/prometheus/client_golang v1.11.0
+	github.com/prometheus/common v0.26.0
+	github.com/sirupsen/logrus v1.8.1
+	github.com/spf13/cobra v1.2.1
+	github.com/spf13/viper v1.8.1
+	github.com/stretchr/testify v1.7.0
 	//github.com/stretchr/testify v1.4.0
-	gopkg.in/yaml.v2 v2.3.0
-	istio.io/api v0.0.0-20180824201241-76349c53b87f
-	k8s.io/api v0.18.6
-	k8s.io/apiextensions-apiserver v0.18.6
-	k8s.io/apimachinery v0.18.6
+	gopkg.in/yaml.v2 v2.4.0
+	istio.io/api v0.0.0-20200715212100-dbf5277541ef
+	k8s.io/api v0.21.4
+	k8s.io/apiextensions-apiserver v0.21.4
+	k8s.io/apimachinery v0.21.4
 	k8s.io/cli-runtime v0.18.5
-	k8s.io/client-go v0.18.6
-	k8s.io/kube-openapi v0.0.0-20200410145947-61e04a5be9a6
-	k8s.io/kubectl v0.0.0
-	sigs.k8s.io/controller-runtime v0.6.5
+	k8s.io/client-go v11.0.1-0.20190805182717-6502b5e7b1b5+incompatible
+	k8s.io/code-generator v0.18.8
+	k8s.io/kube-openapi v0.0.0-20210305001622-591a79e4bda7
+	k8s.io/kubectl v0.17.2
+	knative.dev/pkg v0.0.0-20200922164940-4bf40ad82aab
+	knative.dev/serving v0.18.0
+	sigs.k8s.io/controller-runtime v0.7.0
 )
 
 replace (
-	golang.org/x/sys => golang.org/x/sys v0.0.0-20200826173525-f9321e4c35a6
+	github.com/go-openapi/spec => github.com/go-openapi/spec v0.19.4
+	github.com/swaggo/swag => github.com/swaggo/swag v1.6.7
+	golang.org/x/sys => golang.org/x/sys v0.0.0-20211013075003-97ac67df715c
 	k8s.io/api => k8s.io/api v0.18.5
 	k8s.io/apiextensions-apiserver => k8s.io/apiextensions-apiserver v0.18.5
 	k8s.io/apimachinery => k8s.io/apimachinery v0.18.5
@@ -44,6 +50,7 @@ replace (
 	k8s.io/csi-translation-lib => k8s.io/csi-translation-lib v0.18.5
 	k8s.io/kube-aggregator => k8s.io/kube-aggregator v0.18.5
 	k8s.io/kube-controller-manager => k8s.io/kube-controller-manager v0.18.5
+	k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20200410145947-bcb3869e6f29
 	k8s.io/kube-proxy => k8s.io/kube-proxy v0.18.5
 	k8s.io/kube-scheduler => k8s.io/kube-scheduler v0.18.5
 	k8s.io/kubectl => k8s.io/kubectl v0.18.5