Merge branch 'kubeflow:master' into feature/success_failure_policy

.github/workflows/integration-tests.yaml

@@ -8,16 +8,17 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        kubernetes-version: ["v1.21.12", "v1.22.9", "v1.23.6"]
+        kubernetes-version: ["v1.22.15", "v1.23.12", "v1.24.6"]
     steps:
       - name: Checkout
         uses: actions/checkout@v2
 
       - name: Create k8s Kind Cluster
-        uses: helm/kind-action@v1.2.0
+        uses: helm/kind-action@v1.3.0
         with:
           node_image: kindest/node:${{ matrix.kubernetes-version }}
           cluster_name: training-operator-cluster
+          kubectl_version: ${{ matrix.kubernetes-version }}
 
       - name: Build training-operator
         run: |
@@ -35,4 +36,4 @@ jobs:
       - name: Run tests
         run: |
           pip install pytest
-          python3 -m pip install -r sdk/python/requirements.txt; pytest sdk/python/test --log-cli-level=info
+          python3 -m pip install -e sdk/python; pytest sdk/python/test --log-cli-level=info

.github/workflows/test-go.yaml

@@ -21,9 +21,9 @@ jobs:
           path: ${{ env.GOPATH }}/src/github.com/kubeflow/training-operator
 
       - name: Setup Go
-        uses: actions/setup-go@v2
+        uses: actions/setup-go@v3
         with:
-          go-version: 1.17.2
+          go-version-file: ${{ env.GOPATH }}/src/github.com/kubeflow/training-operator/go.mod
 
       - name: Check Go modules
         run: |

.github/workflows/unittests.yaml

@@ -17,7 +17,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        kubernetes-version: ["1.21.4", "1.22.1", "1.23.5"]
+        kubernetes-version: ["1.22.1", "1.23.5", "1.24.2"]
 
     steps:
       - name: Check out code
@@ -26,9 +26,9 @@ jobs:
           path: ${{ env.GOPATH }}/src/github.com/kubeflow/training-operator
 
       - name: Setup Go
-        uses: actions/setup-go@v2
+        uses: actions/setup-go@v3
         with:
-          go-version: 1.17.2
+          go-version-file: ${{ env.GOPATH }}/src/github.com/kubeflow/training-operator/go.mod
 
       - name: Run Go test
         run: |

Makefile

@@ -58,12 +58,12 @@ vet: ## Run go vet against code.
 GOLANGCI_LINT=$(shell which golangci-lint)
 golangci-lint:
 ifeq ($(GOLANGCI_LINT),)
-	curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh | sh -s -- -b $(shell go env GOPATH)/bin v1.42.1
+	curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh | sh -s -- -b $(shell go env GOPATH)/bin v1.49.0
 	$(info golangci-lint has been installed)
 endif
-	golangci-lint run --timeout 5m ./...
+	golangci-lint run --timeout 5m --go 1.19 ./...
 
-ENVTEST_K8S_VERSION ?= 1.22
+ENVTEST_K8S_VERSION ?= 1.24
 HAS_SETUP_ENVTEST := $(shell command -v setup-envtest;)
 
 testall: manifests generate fmt vet golangci-lint test ## Run tests.
@@ -105,25 +105,12 @@ deploy: manifests kustomize ## Deploy controller to the K8s cluster specified in
 undeploy: ## Undeploy controller from the K8s cluster specified in ~/.kube/config.
 	$(KUSTOMIZE) build manifests/overlays/standalone | kubectl delete -f -
 
+PROJECT_DIR := $(shell dirname $(abspath $(lastword $(MAKEFILE_LIST))))
 
 CONTROLLER_GEN = $(shell pwd)/bin/controller-gen
 controller-gen: ## Download controller-gen locally if necessary.
-	$(call go-get-tool,$(CONTROLLER_GEN),sigs.k8s.io/controller-tools/cmd/[email protected])
+	GOBIN=$(PROJECT_DIR)/bin go install sigs.k8s.io/controller-tools/cmd/[email protected]
 
 KUSTOMIZE = $(shell pwd)/bin/kustomize
 kustomize: ## Download kustomize locally if necessary.
-	$(call go-get-tool,$(KUSTOMIZE),sigs.k8s.io/kustomize/kustomize/[email protected])
-
-# go-get-tool will 'go get' any package $2 and install it to $1.
-PROJECT_DIR := $(shell dirname $(abspath $(lastword $(MAKEFILE_LIST))))
-define go-get-tool
-@[ -f $(1) ] || { \
-set -e ;\
-TMP_DIR=$$(mktemp -d) ;\
-cd $$TMP_DIR ;\
-go mod init tmp ;\
-echo "Downloading $(2)" ;\
-GOBIN=$(PROJECT_DIR)/bin go get $(2) ;\
-rm -rf $$TMP_DIR ;\
-}
-endef
+	GOBIN=$(PROJECT_DIR)/bin go install sigs.k8s.io/kustomize/kustomize/[email protected]

README.md

@@ -38,7 +38,7 @@ kubectl apply -k "github.com/kubeflow/training-operator/manifests/overlays/stand
 ### Stable Release
 
 ```bash
-kubectl apply -k "github.com/kubeflow/training-operator/manifests/overlays/standalone?ref=v1.3.0"
+kubectl apply -k "github.com/kubeflow/training-operator/manifests/overlays/standalone?ref=v1.5.0"
 ```
 
 ### TensorFlow Release Only

build/images/training-operator/Dockerfile

@@ -1,5 +1,5 @@
 # Build the manager binary
-FROM golang:1.17 as builder
+FROM golang:1.19 as builder
 
 WORKDIR /workspace
 # Copy the Go Modules manifests
@@ -13,7 +13,7 @@ RUN go mod download
 COPY . .
 
 # Build
-RUN CGO_ENABLED=0 GOOS=linux GOARCH=amd64 GO111MODULE=on go build -a -o manager cmd/training-operator.v1/main.go
+RUN CGO_ENABLED=0 GOOS=linux GO111MODULE=on go build -a -o manager cmd/training-operator.v1/main.go
 
 # Use distroless as minimal base image to package the manager binary
 # Refer to https://github.com/GoogleContainerTools/distroless for more details

cmd/training-operator.v1/main.go

@@ -29,6 +29,7 @@ import (
 	ctrl "sigs.k8s.io/controller-runtime"
 	"sigs.k8s.io/controller-runtime/pkg/healthz"
 	"sigs.k8s.io/controller-runtime/pkg/log/zap"
+	"volcano.sh/apis/pkg/apis/scheduling/v1beta1"
 
 	commonutil "github.com/kubeflow/common/pkg/util"
 	kubeflowv1 "github.com/kubeflow/training-operator/pkg/apis/kubeflow.org/v1"
@@ -45,12 +46,14 @@ var (
 func init() {
 	utilruntime.Must(clientgoscheme.AddToScheme(scheme))
 	utilruntime.Must(kubeflowv1.AddToScheme(scheme))
+	utilruntime.Must(v1beta1.AddToScheme(scheme))
 	//+kubebuilder:scaffold:scheme
 }
 
 func main() {
 	var metricsAddr string
 	var enableLeaderElection bool
+	var leaderElectionID string
 	var probeAddr string
 	var enabledSchemes controllerv1.EnabledSchemes
 	var enableGangScheduling bool
@@ -62,6 +65,7 @@ func main() {
 	flag.BoolVar(&enableLeaderElection, "leader-elect", false,
 		"Enable leader election for controller manager. "+
 			"Enabling this will ensure there is only one active controller manager.")
+	flag.StringVar(&leaderElectionID, "leader-election-id", "1ca428e5.training-operator.kubeflow.org", "The ID for leader election.")
 	flag.Var(&enabledSchemes, "enable-scheme", "Enable scheme(s) as --enable-scheme=tfjob --enable-scheme=pytorchjob, case insensitive."+
 		" Now supporting TFJob, PyTorchJob, MXNetJob, XGBoostJob. By default, all supported schemes will be enabled.")
 	flag.BoolVar(&enableGangScheduling, "enable-gang-scheduling", false, "Set true to enable gang scheduling")
@@ -96,7 +100,7 @@ func main() {
 		Port:                   monitoringPort,
 		HealthProbeBindAddress: probeAddr,
 		LeaderElection:         enableLeaderElection,
-		LeaderElectionID:       "1ca428e5.",
+		LeaderElectionID:       leaderElectionID,
 		Namespace:              namespace,
 	})
 	if err != nil {

go.mod

@@ -1,6 +1,6 @@
 module github.com/kubeflow/training-operator
 
-go 1.17
+go 1.19
 
 require (
 	github.com/go-logr/logr v1.2.3