e2e: Use helm to deploy on k8s cluster

Using `helm` to deploy the operator in vanilla k8s CI lanes helps avoid error in helm charts changes. Instead of adding a new helm e2e lane (as it would increase the CI load), we can keep the already in place lanes as: - OpenShift: use the `hack/deploy-setup.sh` script - K8s: use helm chart Signed-off-by: Andrea Panattoni <[email protected]>
k8snetworkplumbingwg · May 28, 2024 · 8b1a4d5 · 8b1a4d5
1 parent 8d32f42
commit 8b1a4d5
Show file tree

Hide file tree

Showing 2 changed files with 32 additions and 10 deletions.
diff --git a/Makefile b/Makefile
@@ -172,6 +172,16 @@ skopeo:
 fakechroot:
 	if ! which fakechroot; then if [ -f /etc/redhat-release ]; then dnf -y install fakechroot; elif [ -f /etc/lsb-release ]; then sudo apt-get -y update; sudo apt-get -y install fakechroot; fi; fi
 
+$(BIN_DIR)/get_helm.sh:
+	mkdir -p $(BIN_DIR)
+	curl -fsSL -o $(BIN_DIR)/get_helm.sh https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3
+	chmod 700 $(BIN_DIR)/get_helm.sh
+
+$(BIN_DIR)/helm: $(BIN_DIR)/get_helm.sh
+	HELM_INSTALL_DIR=$(BIN_DIR) $(BIN_DIR)/get_helm.sh
+
+helm: $(BIN_DIR)/helm
+
 deploy-setup: export ADMISSION_CONTROLLERS_ENABLED?=false
 deploy-setup: skopeo install
 	hack/deploy-setup.sh $(NAMESPACE)

diff --git a/hack/run-e2e-conformance-virtual-cluster.sh b/hack/run-e2e-conformance-virtual-cluster.sh
@@ -361,19 +361,37 @@ do
 done
 
 
+source hack/env.sh
+
 export ADMISSION_CONTROLLERS_ENABLED=true
 export ADMISSION_CONTROLLERS_CERTIFICATES_CERT_MANAGER_ENABLED=true
 export SKIP_VAR_SET=""
 export NAMESPACE="sriov-network-operator"
 export OPERATOR_NAMESPACE="sriov-network-operator"
 export CNI_BIN_PATH=/opt/cni/bin
 export OPERATOR_EXEC=kubectl
-export CLUSTER_TYPE=kubernetes
-export DEV_MODE=TRUE
+#export CLUSTER_TYPE=kubernetes
+#export DEV_MODE=TRUE
 export CLUSTER_HAS_EMULATED_PF=TRUE
 
-echo "## deploy namespace"
-envsubst< $root/deploy/namespace.yaml | ${OPERATOR_EXEC} apply -f -
+
+HELM_VALUES_OPTS="\
+  --set operator.images.operator=${SRIOV_NETWORK_OPERATOR_IMAGE} \
+  --set operator.images.sriovConfigDaemon=${SRIOV_NETWORK_CONFIG_DAEMON_IMAGE} \
+  --set operator.images.sriovCni=${SRIOV_CNI_IMAGE} \
+  --set operator.images.sriovDevicePlugin=${SRIOV_DEVICE_PLUGIN_IMAGE} \
+  --set operator.images.resourcesInjector=${NETWORK_RESOURCES_INJECTOR_IMAGE} \
+  --set operator.images.webhook=${SRIOV_NETWORK_WEBHOOK_IMAGE} \
+  --set operator.admissionControllers.enabled=${ADMISSION_CONTROLLERS_ENABLED} \
+  --set operator.admissionControllers.certificates.certManager.enabled=${ADMISSION_CONTROLLERS_CERTIFICATES_CERT_MANAGER_ENABLED} \
+  --set operator.sriovOperatorConfig.deploy=true"
+
+
+make helm
+./bin/helm  install -n ${NAMESPACE} --create-namespace \
+  $HELM_VALUES_OPTS \
+  --wait sriov-network-operator ./
+
 
 echo "## create certificates for webhook"
 cat <<EOF | kubectl apply -f -
@@ -418,12 +436,6 @@ spec:
 EOF
 
 
-echo "## apply CRDs"
-kubectl apply -k $root/config/crd
-
-echo "## deploying SRIOV Network Operator"
-hack/deploy-setup.sh $NAMESPACE
-
 echo "## wait for sriov operator to be ready"
 hack/deploy-wait.sh