chore: Update MLServer to v1.3.2 (kserve#355)

Update MLServer image for the default serving runtime from 0.5.2 to 1.3.2 Depends on kserve/modelmesh-runtime-adapter#45 Closes kserve#357 Signed-off-by: Rafael Vasquez <[email protected]> Signed-off-by: Luis Delossantos <[email protected]>
lgdeloss · Jun 5, 2023 · 5bf1ec2 · 5bf1ec2
1 parent d26ca20
commit 5bf1ec2
Show file tree

Hide file tree

Showing 17 changed files with 54 additions and 54 deletions.
diff --git a/.github/workflows/fvt.yml b/.github/workflows/fvt.yml
@@ -71,11 +71,11 @@ jobs:
           sed -i 's/newTag:.*$/newTag: '"${{ env.IMAGE_TAG }}"'/' config/manager/kustomization.yaml
           sed -i '0,/cpu:.*$/s/cpu:.*$/cpu: 100m/' \
             config/default/config-defaults.yaml \
-            config/runtimes/mlserver-0.x.yaml \
+            config/runtimes/mlserver-1.x.yaml \
             config/runtimes/triton-2.x.yaml \
             config/runtimes/ovms-1.x.yaml
           sed -i 's/memory:.*$/memory: 512Mi/g' \
-            config/runtimes/mlserver-0.x.yaml \
+            config/runtimes/mlserver-1.x.yaml \
             config/runtimes/triton-2.x.yaml \
             config/runtimes/ovms-1.x.yaml
           sed -i 's/maxSurge:.*$/maxSurge: 0/' config/internal/base/deployment.yaml.tmpl
@@ -109,7 +109,7 @@ jobs:
         run: |
           eval $(minikube -p minikube docker-env)
           docker pull nvcr.io/nvidia/tritonserver:21.06.1-py3
-          docker pull seldonio/mlserver:0.5.2
+          docker pull seldonio/mlserver:1.3.2
           docker pull openvino/model_server:2022.2
           # docker pull pytorch/torchserve:0.7.1-cpu
           docker pull kserve/modelmesh-runtime-adapter

diff --git a/config/runtimes/kustomization.yaml b/config/runtimes/kustomization.yaml
@@ -13,7 +13,7 @@
 # limitations under the License.
 resources:
   - triton-2.x.yaml
-  - mlserver-0.x.yaml
+  - mlserver-1.x.yaml
   - ovms-1.x.yaml
   - torchserve-0.x.yaml
 
@@ -22,9 +22,9 @@ images:
     newName: nvcr.io/nvidia/tritonserver
     newTag: "21.06.1-py3"
 
-  - name: mlserver-0
+  - name: mlserver-1
     newName: seldonio/mlserver
-    newTag: "0.5.2"
+    newTag: "1.3.2"
 
   - name: ovms-1
     newName: openvino/model_server

diff --git a/config/runtimes/mlserver-0.x.yaml → config/runtimes/mlserver-1.x.yaml b/config/runtimes/mlserver-0.x.yaml → config/runtimes/mlserver-1.x.yaml
@@ -14,9 +14,9 @@
 apiVersion: serving.kserve.io/v1alpha1
 kind: ClusterServingRuntime
 metadata:
-  name: mlserver-0.x
+  name: mlserver-1.x
   labels:
-    name: modelmesh-serving-mlserver-0.x-SR
+    name: modelmesh-serving-mlserver-1.x-SR
 spec:
   supportedModelFormats:
     - name: sklearn
@@ -38,7 +38,7 @@ spec:
 
   containers:
     - name: mlserver
-      image: mlserver-0:replace
+      image: mlserver-1:replace
       env:
         - name: MLSERVER_MODELS_DIR
           value: "/models/_mlserver_models/"

diff --git a/config/samples/servingruntime_pullerless.yaml b/config/samples/servingruntime_pullerless.yaml
@@ -12,7 +12,7 @@ spec:
             resourceFieldRef:
               containerName: modelserver
               resource: requests.memory
-      image: seldonio/mlserver:0.3.2
+      image: seldonio/mlserver:1.3.2
       name: modelserver
       resources:
         requests:

diff --git a/controllers/servingruntime_controller_test.go b/controllers/servingruntime_controller_test.go
@@ -4,7 +4,7 @@
 // you may not use this file except in compliance with the License.
 // You may obtain a copy of the License at
 //
-//     http://www.apache.org/licenses/LICENSE-2.0
+//	http://www.apache.org/licenses/LICENSE-2.0
 //
 // Unless required by applicable law or agreed to in writing, software
 // distributed under the License is distributed on an "AS IS" BASIS,
@@ -53,7 +53,7 @@ func waitForAndGetRuntimeDeployment(runtimeName string) *appsv1.Deployment {
 
 var _ = Describe("Sample Runtime", func() {
 	samplesToTest := []string{
-		"config/runtimes/mlserver-0.x.yaml",
+		"config/runtimes/mlserver-1.x.yaml",
 		"config/runtimes/triton-2.x.yaml",
 		"config/runtimes/ovms-1.x.yaml",
 		"config/runtimes/torchserve-0.x.yaml",
@@ -98,7 +98,7 @@ var _ = Describe("Prometheus metrics configuration", func() {
 		reconcilerConfig.Metrics.Enabled = true
 
 		By("create a sample runtime")
-		m, err = mf.ManifestFrom(mf.Path("../config/runtimes/mlserver-0.x.yaml"))
+		m, err = mf.ManifestFrom(mf.Path("../config/runtimes/mlserver-1.x.yaml"))
 		m.Client = mfc.NewClient(k8sClient)
 		Expect(err).ToNot(HaveOccurred())
 		m, err = m.Transform(convertToServingRuntime, mf.InjectNamespace(namespace))
@@ -177,7 +177,7 @@ var _ = Describe("REST Proxy configuration", func() {
 		reconcilerConfig.RESTProxy.Enabled = true
 
 		By("create a sample runtime")
-		m, err = mf.ManifestFrom(mf.Path("../config/runtimes/mlserver-0.x.yaml"))
+		m, err = mf.ManifestFrom(mf.Path("../config/runtimes/mlserver-1.x.yaml"))
 		m.Client = mfc.NewClient(k8sClient)
 		Expect(err).ToNot(HaveOccurred())
 		m, err = m.Transform(convertToServingRuntime, mf.InjectNamespace(namespace))
@@ -204,7 +204,7 @@ var _ = Describe("Add Payload Processor", func() {
 		resetToPayloadConfig(false)
 
 		By("create a sample runtime")
-		m, err = mf.ManifestFrom(mf.Path("../config/runtimes/mlserver-0.x.yaml"))
+		m, err = mf.ManifestFrom(mf.Path("../config/runtimes/mlserver-1.x.yaml"))
 		m.Client = mfc.NewClient(k8sClient)
 		Expect(err).ToNot(HaveOccurred())
 		m, err = m.Transform(convertToServingRuntime, mf.InjectNamespace(namespace))

diff --git a/controllers/testdata/servingruntime_controller.golden b/controllers/testdata/servingruntime_controller.golden
@@ -12,7 +12,7 @@ spec:
   selector:
     matchLabels:
       modelmesh-service: modelmesh-serving
-      name: modelmesh-serving-mlserver-0.x
+      name: modelmesh-serving-mlserver-1.x
   strategy:
     rollingUpdate:
       maxSurge: 75%
@@ -26,7 +26,7 @@ spec:
         app.kubernetes.io/managed-by: modelmesh-controller
         app.kubernetes.io/name: modelmesh-controller
         modelmesh-service: modelmesh-serving
-        name: modelmesh-serving-mlserver-0.x
+        name: modelmesh-serving-mlserver-1.x
     spec:
       affinity:
         nodeAffinity:
@@ -105,7 +105,7 @@ spec:
           value: 127.0.0.1
         - name: MLSERVER_GRPC_MAX_MESSAGE_LENGTH
           value: "-1"
-        image: mlserver-0:replace
+        image: mlserver-1:replace
         imagePullPolicy: IfNotPresent
         lifecycle:
           preStop:
@@ -169,7 +169,7 @@ spec:
         - name: MM_PAYLOAD_PROCESSORS
           value: example:8080/consumer/kserve/v2 example2:8080/consumer/kserve/v2
         - name: MM_LABELS
-          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,rt:mlserver-0.x
+          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,rt:mlserver-1.x
         - name: MM_TYPE_CONSTRAINTS_PATH
           value: /etc/watson/mmesh/config/type_constraints
         - name: MM_DATAPLANE_CONFIG_PATH
@@ -261,7 +261,7 @@ spec:
   selector:
     matchLabels:
       modelmesh-service: modelmesh-serving
-      name: modelmesh-serving-mlserver-0.x
+      name: modelmesh-serving-mlserver-1.x
   strategy:
     rollingUpdate:
       maxSurge: 75%
@@ -280,7 +280,7 @@ spec:
         app.kubernetes.io/managed-by: modelmesh-controller
         app.kubernetes.io/name: modelmesh-controller
         modelmesh-service: modelmesh-serving
-        name: modelmesh-serving-mlserver-0.x
+        name: modelmesh-serving-mlserver-1.x
     spec:
       affinity:
         nodeAffinity:
@@ -359,7 +359,7 @@ spec:
           value: 127.0.0.1
         - name: MLSERVER_GRPC_MAX_MESSAGE_LENGTH
           value: "-1"
-        image: mlserver-0:replace
+        image: mlserver-1:replace
         imagePullPolicy: IfNotPresent
         lifecycle:
           preStop:
@@ -421,7 +421,7 @@ spec:
         - name: MM_DEFAULT_VMODEL_OWNER
           value: ksp
         - name: MM_LABELS
-          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,rt:mlserver-0.x
+          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,rt:mlserver-1.x
         - name: MM_TYPE_CONSTRAINTS_PATH
           value: /etc/watson/mmesh/config/type_constraints
         - name: MM_DATAPLANE_CONFIG_PATH
@@ -516,7 +516,7 @@ spec:
   selector:
     matchLabels:
       modelmesh-service: modelmesh-serving
-      name: modelmesh-serving-mlserver-0.x
+      name: modelmesh-serving-mlserver-1.x
   strategy:
     rollingUpdate:
       maxSurge: 75%
@@ -530,7 +530,7 @@ spec:
         app.kubernetes.io/managed-by: modelmesh-controller
         app.kubernetes.io/name: modelmesh-controller
         modelmesh-service: modelmesh-serving
-        name: modelmesh-serving-mlserver-0.x
+        name: modelmesh-serving-mlserver-1.x
     spec:
       affinity:
         nodeAffinity:
@@ -583,7 +583,7 @@ spec:
           value: 127.0.0.1
         - name: MLSERVER_GRPC_MAX_MESSAGE_LENGTH
           value: "-1"
-        image: mlserver-0:replace
+        image: mlserver-1:replace
         imagePullPolicy: IfNotPresent
         lifecycle:
           preStop:
@@ -696,7 +696,7 @@ spec:
         - name: MM_DEFAULT_VMODEL_OWNER
           value: ksp
         - name: MM_LABELS
-          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,pv:v2,rt:mlserver-0.x
+          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,pv:v2,rt:mlserver-1.x
         - name: MM_TYPE_CONSTRAINTS_PATH
           value: /etc/watson/mmesh/config/type_constraints
         - name: MM_DATAPLANE_CONFIG_PATH
@@ -776,7 +776,7 @@ spec:
           secretName: secret
 status: {}
 '''
-"Sample Runtime config/runtimes/mlserver-0.x.yaml should be a valid runtime specification" = '''
+"Sample Runtime config/runtimes/mlserver-1.x.yaml should be a valid runtime specification" = '''
 apiVersion: apps/v1
 kind: Deployment
 metadata:
@@ -788,7 +788,7 @@ spec:
   selector:
     matchLabels:
       modelmesh-service: modelmesh-serving
-      name: modelmesh-serving-mlserver-0.x
+      name: modelmesh-serving-mlserver-1.x
   strategy:
     rollingUpdate:
       maxSurge: 75%
@@ -802,7 +802,7 @@ spec:
         app.kubernetes.io/managed-by: modelmesh-controller
         app.kubernetes.io/name: modelmesh-controller
         modelmesh-service: modelmesh-serving
-        name: modelmesh-serving-mlserver-0.x
+        name: modelmesh-serving-mlserver-1.x
     spec:
       affinity:
         nodeAffinity:
@@ -881,7 +881,7 @@ spec:
           value: 127.0.0.1
         - name: MLSERVER_GRPC_MAX_MESSAGE_LENGTH
           value: "-1"
-        image: mlserver-0:replace
+        image: mlserver-1:replace
         imagePullPolicy: IfNotPresent
         lifecycle:
           preStop:
@@ -943,7 +943,7 @@ spec:
         - name: MM_DEFAULT_VMODEL_OWNER
           value: ksp
         - name: MM_LABELS
-          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,rt:mlserver-0.x
+          value: mt:lightgbm,mt:lightgbm:3,mt:sklearn,mt:sklearn:0,mt:xgboost,mt:xgboost:1,pv:grpc-v2,rt:mlserver-1.x
         - name: MM_TYPE_CONSTRAINTS_PATH
           value: /etc/watson/mmesh/config/type_constraints
         - name: MM_DATAPLANE_CONFIG_PATH
@@ -1809,7 +1809,7 @@ spec:
               containerName: modelserver
               divisor: "0"
               resource: requests.memory
-        image: seldonio/mlserver:0.3.2
+        image: seldonio/mlserver:1.3.2
         imagePullPolicy: IfNotPresent
         lifecycle:
           preStop:

diff --git a/docs/install/README.md b/docs/install/README.md
@@ -64,7 +64,7 @@ The deployed footprint can be significantly reduced in the following ways:
 
 ```shell
 > kubectl edit servingruntime triton-2.x
-> kubectl edit servingruntime mlserver-0.x
+> kubectl edit servingruntime mlserver-1.x
 > kubectl edit servingruntime ovms-1.x
 ```
 

diff --git a/docs/model-formats/advanced-configuration.md b/docs/model-formats/advanced-configuration.md
@@ -78,8 +78,8 @@ To pass runtime specific configuration through to MLServer, include a non-empty
 └── <model-data>
 ```
 
-For details on the specification of this configuration, refer to the definition of
-[MLServer's ModelSettings class](https://github.com/SeldonIO/MLServer/blob/0.3.2/mlserver/settings.py#L49).
+For details on the specification of this configuration, refer to
+[MLServer's Model Settings documentation](https://github.com/SeldonIO/MLServer/blob/1.3.2/docs/reference/model-settings.md).
 
 ---
 

diff --git a/docs/monitoring.md b/docs/monitoring.md
@@ -73,7 +73,7 @@ If the ModelMesh Serving metric(s) are missing in the monitoring UIs:
 
 - Check if the following annotations are configured in the Serving Runtime deployment:
 
-        kubectl describe deployment modelmesh-serving-mlserver-0.x -n $NAMESPACE | grep "prometheus.io"
+        kubectl describe deployment modelmesh-serving-mlserver-1.x -n $NAMESPACE | grep "prometheus.io"
 
   Expected output:
 

diff --git a/docs/predictors/README.md b/docs/predictors/README.md
@@ -69,8 +69,8 @@ Once the `InferenceService` is created, mlserver runtime pods are automatically
 $ kubectl get pods
 
 NAME                                               READY   STATUS              RESTARTS   AGE
-modelmesh-serving-mlserver-0.x-658b7dd689-46nwm    0/3     ContainerCreating   0          2s
-modelmesh-serving-mlserver-0.x-658b7dd689-46nwm    0/3     ContainerCreating   0          2s
+modelmesh-serving-mlserver-1.x-658b7dd689-46nwm    0/3     ContainerCreating   0          2s
+modelmesh-serving-mlserver-1.x-658b7dd689-46nwm    0/3     ContainerCreating   0          2s
 modelmesh-controller-568c45b959-nl88c              1/1     Running             0          11m
 ```
 

diff --git a/docs/quickstart.md b/docs/quickstart.md
@@ -51,7 +51,7 @@ Check that the `ServingRuntime`s are available:
 kubectl get servingruntimes
 
 NAME           DISABLED   MODELTYPE    CONTAINERS   AGE
-mlserver-0.x              sklearn      mlserver     5m
+mlserver-1.x              sklearn      mlserver     5m
 ovms-1.x                  openvino_ir  ovms         5m
 torchserve-0.x            pytorch-mar  torchserve   5m
 triton-2.x                tensorflow   triton       5m
@@ -63,7 +63,7 @@ are:
 
 | ServingRuntime | Supported Frameworks                |
 | -------------- | ----------------------------------- |
-| mlserver-0.x   | sklearn, xgboost, lightgbm          |
+| mlserver-1.x   | sklearn, xgboost, lightgbm          |
 | ovms-1.x       | openvino_ir, onnx                   |
 | torchserve-0.x | pytorch-mar                         |
 | triton-2.x     | tensorflow, pytorch, onnx, tensorrt |
@@ -134,8 +134,8 @@ Eventually, you should see the `ServingRuntime` pods that will hold the SKLearn
 kubectl get pods
 
 ...
-modelmesh-serving-mlserver-0.x-7db675f677-twrwd   3/3     Running   0          2m
-modelmesh-serving-mlserver-0.x-7db675f677-xvd8q   3/3     Running   0          2m
+modelmesh-serving-mlserver-1.x-7db675f677-twrwd   3/3     Running   0          2m
+modelmesh-serving-mlserver-1.x-7db675f677-xvd8q   3/3     Running   0          2m
 ```
 
 Then, checking on the `InferenceService` again, you should see that the one we deployed is now ready with a provided URL:

diff --git a/fvt/fvtclient.go b/fvt/fvtclient.go
@@ -4,7 +4,7 @@
 // you may not use this file except in compliance with the License.
 // You may obtain a copy of the License at
 //
-//     http://www.apache.org/licenses/LICENSE-2.0
+//	http://www.apache.org/licenses/LICENSE-2.0
 //
 // Unless required by applicable law or agreed to in writing, software
 // distributed under the License is distributed on an "AS IS" BASIS,

diff --git a/fvt/hpa/hpa_test.go b/fvt/hpa/hpa_test.go
@@ -4,7 +4,7 @@
 // you may not use this file except in compliance with the License.
 // You may obtain a copy of the License at
 //
-//     http://www.apache.org/licenses/LICENSE-2.0
+//	http://www.apache.org/licenses/LICENSE-2.0
 //
 // Unless required by applicable law or agreed to in writing, software
 // distributed under the License is distributed on an "AS IS" BASIS,
@@ -31,7 +31,7 @@ var _ = Describe("Scaling of runtime deployments with HPA Autoscaler", Ordered,
 	// constants
 	testPredictorObject := NewPredictorForFVT("mlserver-sklearn-predictor.yaml")
 	// runtime expected to serve the test predictor
-	expectedRuntimeName := "mlserver-0.x"
+	expectedRuntimeName := "mlserver-1.x"
 
 	// checkDeploymentState returns the replicas value for the expected runtime
 	// and expects others to be scaled to zero