Update Tiltfile with AKS VNet peering and deletion logic

- update aks-as-mgmt scripts with VNet creation and all clusters deletion - update default template - update default aad - update azure-bastion templates - update azure-cni-v1 templates - update edgezone templates - update ephemeral templates - update private templates - update flatcar templates - update nvdia-gpu templates
kubernetes-sigs · Nov 19, 2024 · 3025950 · 3025950
1 parent 6bc2cab
commit 3025950
Show file tree

Hide file tree

Showing 48 changed files with 794 additions and 87 deletions.
diff --git a/Tiltfile b/Tiltfile
@@ -23,8 +23,8 @@ settings = {
     "capi_version": "v1.8.5",
     "caaph_version": "v0.2.5",
     "cert_manager_version": "v1.16.1",
-    "kubernetes_version": "v1.28.3",
-    "aks_kubernetes_version": "v1.28.3",
+    "kubernetes_version": "v1.28.15",
+    "aks_kubernetes_version": "v1.28.15",
     "flatcar_version": "3374.2.1",
     "azure_location": "eastus",
     "control_plane_machine_count": "1",
@@ -212,10 +212,10 @@ def capz():
     yaml = str(kustomizesub("./hack/observability"))  # build an observable kind deployment by default
 
     # add extra_args if they are defined
-    if settings.get("extra_args"):
-        azure_extra_args = settings.get("extra_args").get("azure")
+    if settings.get("container_args"):
+        capz_container_args = settings.get("container_args").get("capz-controller-manager")
         yaml_dict = decode_yaml_stream(yaml)
-        append_arg_for_container_in_deployment(yaml_dict, "capz-controller-manager", "capz-system", "cluster-api-azure-controller", azure_extra_args)
+        append_arg_for_container_in_deployment(yaml_dict, "capz-controller-manager", "capz-system", "cluster-api-azure-controller", capz_container_args)
         yaml = str(encode_yaml_stream(yaml_dict))
         yaml = fixup_yaml_empty_arrays(yaml)
 
@@ -317,9 +317,14 @@ def flavors():
     for template in template_list:
         deploy_worker_templates(template, substitutions)
 
+    delete_all_workload_clusters = kubectl_cmd + " delete clusters --all --wait=false"
+
+    if "aks" in settings.get("kustomize_substitutions", {}).get("MGMT_CLUSTER_NAME", ""):
+        delete_all_workload_clusters += clear_aks_vnet_peerings()
+
     local_resource(
         name = "delete-all-workload-clusters",
-        cmd = kubectl_cmd + " delete clusters --all --wait=false",
+        cmd = ["sh", "-ec", delete_all_workload_clusters],
         auto_init = False,
         trigger_mode = TRIGGER_MODE_MANUAL,
         labels = ["flavors"],
@@ -382,17 +387,29 @@ def deploy_worker_templates(template, substitutions):
 
     yaml = shlex.quote(yaml)
     flavor_name = os.path.basename(flavor)
-    flavor_cmd = "RANDOM=$(bash -c 'echo $RANDOM'); export CLUSTER_NAME=" + flavor.replace("windows", "win") + "-$RANDOM; make generate-flavors; echo " + yaml + "> ./.tiltbuild/" + flavor + "; cat ./.tiltbuild/" + flavor + " | " + envsubst_cmd + " | " + kubectl_cmd + " apply -f -; echo \"Cluster \'$CLUSTER_NAME\' created, don't forget to delete\""
+    flavor_cmd = "RANDOM=$(bash -c 'echo $RANDOM')"
+    flavor_cmd += "; export CLUSTER_NAME=" + flavor.replace("windows", "win") + "-$RANDOM; echo " + yaml + "> ./.tiltbuild/" + flavor + "; cat ./.tiltbuild/" + flavor + " | " + envsubst_cmd + " | " + kubectl_cmd + " apply -f -"
+    flavor_cmd += "; echo \"Cluster \'$CLUSTER_NAME\' created, don't forget to delete\""
 
     # wait for kubeconfig to be available
-    flavor_cmd += "; until " + kubectl_cmd + " get secret ${CLUSTER_NAME}-kubeconfig > /dev/null 2>&1; do sleep 5; done; " + kubectl_cmd + " get secret ${CLUSTER_NAME}-kubeconfig -o jsonpath={.data.value} | base64 --decode > ./${CLUSTER_NAME}.kubeconfig; chmod 600 ./${CLUSTER_NAME}.kubeconfig; until " + kubectl_cmd + " --kubeconfig=./${CLUSTER_NAME}.kubeconfig get nodes > /dev/null 2>&1; do sleep 5; done"
+    flavor_cmd += "; echo \"Waiting for kubeconfig to be available\""
+    flavor_cmd += "; until " + kubectl_cmd + " get secret ${CLUSTER_NAME}-kubeconfig > /dev/null 2>&1; do sleep 5; done"
+    flavor_cmd += "; " + kubectl_cmd + " get secret ${CLUSTER_NAME}-kubeconfig -o jsonpath={.data.value} | base64 --decode > ./${CLUSTER_NAME}.kubeconfig"
+    flavor_cmd += "; chmod 600 ./${CLUSTER_NAME}.kubeconfig"
+    flavor_cmd += "; echo \"Kubeconfig for $CLUSTER_NAME created and saved in the local\""
+    flavor_cmd += "; echo \"Waiting for $CLUSTER_NAME API Server to be accessible\""
+    flavor_cmd += "; until " + kubectl_cmd + " --kubeconfig=./${CLUSTER_NAME}.kubeconfig get nodes > /dev/null 2>&1; do sleep 5; done"
+    flavor_cmd += "; echo \"API Server of $CLUSTER_NAME is accessible\""
 
     # copy the kubeadm configmap to the calico-system namespace.
     # This is a workaround needed for the calico-node-windows daemonset to be able to run in the calico-system namespace.
     if "windows" in flavor_name:
         flavor_cmd += "; until " + kubectl_cmd + " --kubeconfig ./${CLUSTER_NAME}.kubeconfig get configmap kubeadm-config --namespace=kube-system > /dev/null 2>&1; do sleep 5; done"
         flavor_cmd += "; " + kubectl_cmd + " --kubeconfig ./${CLUSTER_NAME}.kubeconfig create namespace calico-system --dry-run=client -o yaml | " + kubectl_cmd + " --kubeconfig ./${CLUSTER_NAME}.kubeconfig apply -f -; " + kubectl_cmd + " --kubeconfig ./${CLUSTER_NAME}.kubeconfig get configmap kubeadm-config --namespace=kube-system -o yaml | sed 's/namespace: kube-system/namespace: calico-system/' | " + kubectl_cmd + " --kubeconfig ./${CLUSTER_NAME}.kubeconfig apply -f -"
 
+    if "aks" in settings.get("kustomize_substitutions", {}).get("MGMT_CLUSTER_NAME", ""):
+        flavor_cmd += peer_vnets()
+
     flavor_cmd += get_addons(flavor_name)
 
     local_resource(
@@ -454,6 +471,63 @@ def waitforsystem():
     local(kubectl_cmd + " wait --for=condition=ready --timeout=300s pod --all -n capi-kubeadm-control-plane-system")
     local(kubectl_cmd + " wait --for=condition=ready --timeout=300s pod --all -n capi-system")
 
+def peer_vnets():
+    # TODO: check for az cli to be installed in local
+    # wait for AKS VNet to be in the state created
+    peering_cmd = "; echo \"--------Peering VNETs--------\""
+    peering_cmd += "; az network vnet wait --resource-group ${AKS_RESOURCE_GROUP} --name ${AKS_MGMT_VNET_NAME} --created --timeout 180"
+    peering_cmd += "; export MGMT_VNET_ID=$(az network vnet show --resource-group ${AKS_RESOURCE_GROUP} --name ${AKS_MGMT_VNET_NAME} --query id --output tsv)"
+    peering_cmd += "; echo \" 1/8 ${AKS_MGMT_VNET_NAME} found \""
+
+    # wait for workload VNet to be created
+    peering_cmd += "; az network vnet wait --resource-group ${CLUSTER_NAME} --name ${CLUSTER_NAME}-vnet --created --timeout 180"
+    peering_cmd += "; export WORKLOAD_VNET_ID=$(az network vnet show --resource-group ${CLUSTER_NAME} --name ${CLUSTER_NAME}-vnet --query id --output tsv)"
+    peering_cmd += "; echo \" 2/8 ${CLUSTER_NAME}-vnet found \""
+
+    # peer mgmt vnet
+    peering_cmd += "; az network vnet peering create --name mgmt-to-${CLUSTER_NAME} --resource-group ${AKS_RESOURCE_GROUP} --vnet-name ${AKS_MGMT_VNET_NAME} --remote-vnet \"${WORKLOAD_VNET_ID}\" --allow-vnet-access true --allow-forwarded-traffic true --only-show-errors --output none"
+    peering_cmd += "; az network vnet peering wait --name mgmt-to-${CLUSTER_NAME} --resource-group ${AKS_RESOURCE_GROUP} --vnet-name ${AKS_MGMT_VNET_NAME} --created --timeout 300 --only-show-errors --output none"
+    peering_cmd += "; echo \" 3/8 mgmt-to-${CLUSTER_NAME} peering created in ${AKS_MGMT_VNET_NAME}\""
+
+    # peer workload vnet
+    peering_cmd += "; az network vnet peering create --name ${CLUSTER_NAME}-to-mgmt --resource-group ${CLUSTER_NAME} --vnet-name ${CLUSTER_NAME}-vnet --remote-vnet \"${MGMT_VNET_ID}\" --allow-vnet-access true --allow-forwarded-traffic true --only-show-errors --output none"
+    peering_cmd += "; az network vnet peering wait --name ${CLUSTER_NAME}-to-mgmt --resource-group ${CLUSTER_NAME} --vnet-name ${CLUSTER_NAME}-vnet --created --timeout 300 --only-show-errors --output none"
+    peering_cmd += "; echo \" 4/8 ${CLUSTER_NAME}-to-mgmt peering created in ${CLUSTER_NAME}-vnet\""
+
+    # create private DNS zone
+    peering_cmd += "; az network private-dns zone create --resource-group ${CLUSTER_NAME} --name ${AZURE_LOCATION}.cloudapp.azure.com --only-show-errors --output none"
+    peering_cmd += "; az network private-dns zone wait --resource-group ${CLUSTER_NAME} --name ${AZURE_LOCATION}.cloudapp.azure.com --created --timeout 300 --only-show-errors --output none"
+    peering_cmd += "; echo \" 5/8 ${AZURE_LOCATION}.cloudapp.azure.com private DNS zone created in ${CLUSTER_NAME}\""
+
+    # link private DNS Zone to workload vnet
+    peering_cmd += "; az network private-dns link vnet create --resource-group ${CLUSTER_NAME} --zone-name ${AZURE_LOCATION}.cloudapp.azure.com --name ${CLUSTER_NAME}-to-mgmt --virtual-network \"${WORKLOAD_VNET_ID}\" --registration-enabled false --only-show-errors --output none"
+    peering_cmd += "; az network private-dns link vnet wait --resource-group ${CLUSTER_NAME} --zone-name ${AZURE_LOCATION}.cloudapp.azure.com --name ${CLUSTER_NAME}-to-mgmt --created --timeout 300 --only-show-errors --output none"
+    peering_cmd += "; echo \" 6/8 workload cluster vnet ${CLUSTER_NAME}-vnet linked with private DNS zone\""
+
+    # link private DNS Zone to mgmt vnet
+    peering_cmd += "; az network private-dns link vnet create --resource-group ${CLUSTER_NAME} --zone-name ${AZURE_LOCATION}.cloudapp.azure.com --name mgmt-to-${CLUSTER_NAME} --virtual-network \"${MGMT_VNET_ID}\" --registration-enabled false --only-show-errors --output none"
+    peering_cmd += "; az network private-dns link vnet wait --resource-group ${CLUSTER_NAME} --zone-name ${AZURE_LOCATION}.cloudapp.azure.com --name mgmt-to-${CLUSTER_NAME} --created --timeout 300 --only-show-errors --output none"
+    peering_cmd += "; echo \" 7/8 management cluster vnet ${AKS_MGMT_VNET_NAME} linked with private DNS zone\""
+
+    # create private DNS zone record
+    # TODO: 10.0.0.100 should be customizable
+    peering_cmd += "; az network private-dns record-set a add-record --resource-group ${CLUSTER_NAME} --zone-name ${AZURE_LOCATION}.cloudapp.azure.com --record-set-name ${CLUSTER_NAME}-${APISERVER_LB_DNS_SUFFIX} --ipv4-address 10.0.0.100 --only-show-errors --output none"
+    peering_cmd += "; echo \" 8/8 ${CLUSTER_NAME}-${APISERVER_LB_DNS_SUFFIX} private DNS zone record created\n\""
+
+    return peering_cmd
+
+def clear_aks_vnet_peerings():
+    delete_peering_cmd = "; echo \"--------Clearing AKS MGMT VNETs Peerings--------\""
+    delete_peering_cmd += "; az network vnet wait --resource-group ${AKS_RESOURCE_GROUP} --name ${AKS_MGMT_VNET_NAME} --created --timeout 180"
+    delete_peering_cmd += "; echo \" ${AKS_MGMT_VNET_NAME} found \""
+
+    # List all peering names and store them in an array
+    delete_peering_cmd += "; PEERING_NAMES=$(az network vnet peering list --resource-group ${AKS_RESOURCE_GROUP} --vnet-name ${AKS_MGMT_VNET_NAME} --query \"[].name\" --output tsv)"
+    delete_peering_cmd += "; for PEERING_NAME in ${PEERING_NAMES[@]}; do echo \"Deleting peering: ${PEERING_NAME}\"; az network vnet peering delete --name ${PEERING_NAME} --resource-group ${AKS_RESOURCE_GROUP} --vnet-name ${AKS_MGMT_VNET_NAME}; done"
+    delete_peering_cmd += "; echo \"All VNETs Peerings deleted in ${AKS_MGMT_VNET_NAME}\""
+
+    return delete_peering_cmd
+
 ##############################
 # Actual work happens here
 ##############################

diff --git a/scripts/aks-as-mgmt.sh b/scripts/aks-as-mgmt.sh
@@ -30,7 +30,7 @@ make --directory="${REPO_ROOT}" "${KUBECTL##*/}" "${AZWI##*/}"
 export MGMT_CLUSTER_NAME="${MGMT_CLUSTER_NAME:-aks-mgmt-capz-${RANDOM_SUFFIX}}" # management cluster name
 export AKS_RESOURCE_GROUP="${AKS_RESOURCE_GROUP:-aks-mgmt-capz-${RANDOM_SUFFIX}}" # resource group name
 export AKS_NODE_RESOURCE_GROUP="node-${AKS_RESOURCE_GROUP}"
-export KUBERNETES_VERSION="${KUBERNETES_VERSION:-v1.30.2}"
+export AKS_MGMT_KUBERNETES_VERSION="${AKS_MGMT_KUBERNETES_VERSION:-v1.30.2}"
 export AZURE_LOCATION="${AZURE_LOCATION:-westus2}"
 export AKS_NODE_VM_SIZE="${AKS_NODE_VM_SIZE:-"Standard_B2s"}"
 export AKS_NODE_COUNT="${AKS_NODE_COUNT:-1}"
@@ -42,6 +42,13 @@ export AZWI_STORAGE_CONTAINER="\$web"
 export SERVICE_ACCOUNT_SIGNING_PUB_FILEPATH="${SERVICE_ACCOUNT_SIGNING_PUB_FILEPATH:-}"
 export SERVICE_ACCOUNT_SIGNING_KEY_FILEPATH="${SERVICE_ACCOUNT_SIGNING_KEY_FILEPATH:-}"
 export REGISTRY="${REGISTRY:-}"
+export AKS_MGMT_VNET_NAME="${AKS_MGMT_VNET_NAME:-"aks-mgmt-vnet-${RANDOM_SUFFIX}"}"
+export AKS_MGMT_VNET_CIDR="${AKS_MGMT_VNET_CIDR:-"20.255.0.0/16"}"
+export AKS_MGMT_SERVICE_CIDR="${AKS_MGMT_SERVICE_CIDR:-"20.255.254.0/24"}"
+export AKS_MGMT_DNS_SERVICE_IP="${AKS_MGMT_DNS_SERVICE_IP:-"20.255.254.100"}"
+export AKS_MGMT_SUBNET_NAME="${AKS_MGMT_SUBNET_NAME:-"aks-mgmt-subnet-${RANDOM_SUFFIX}"}"
+export AKS_MGMT_SUBNET_CIDR="${AKS_MGMT_SUBNET_CIDR:-"20.255.0.0/24"}"
+
 
 export AZURE_SUBSCRIPTION_ID="${AZURE_SUBSCRIPTION_ID:-}"
 export AZURE_CLIENT_ID="${AZURE_CLIENT_ID:-}"
@@ -63,7 +70,7 @@ main() {
   echo "MGMT_CLUSTER_NAME:                    $MGMT_CLUSTER_NAME"
   echo "AKS_RESOURCE_GROUP:                   $AKS_RESOURCE_GROUP"
   echo "AKS_NODE_RESOURCE_GROUP:              $AKS_NODE_RESOURCE_GROUP"
-  echo "KUBERNETES_VERSION:                   $KUBERNETES_VERSION"
+  echo "AKS_MGMT_KUBERNETES_VERSION:          $AKS_MGMT_KUBERNETES_VERSION"
   echo "AZURE_LOCATION:                       $AZURE_LOCATION"
   echo "AKS_NODE_VM_SIZE:                     $AKS_NODE_VM_SIZE"
   echo "AZURE_NODE_MACHINE_TYPE:              $AZURE_NODE_MACHINE_TYPE"
@@ -76,6 +83,12 @@ main() {
   echo "SERVICE_ACCOUNT_SIGNING_KEY_FILEPATH: $SERVICE_ACCOUNT_SIGNING_KEY_FILEPATH"
   echo "REGISTRY:                             $REGISTRY"
   echo "APISERVER_LB_DNS_SUFFIX:              $APISERVER_LB_DNS_SUFFIX"
+  echo "AKS_MGMT_VNET_NAME:                   $AKS_MGMT_VNET_NAME"
+  echo "AKS_MGMT_VNET_CIDR:                   $AKS_MGMT_VNET_CIDR"
+  echo "AKS_MGMT_SERVICE_CIDR:                $AKS_MGMT_SERVICE_CIDR"
+  echo "AKS_MGMT_DNS_SERVICE_IP:              $AKS_MGMT_DNS_SERVICE_IP"
+  echo "AKS_MGMT_SUBNET_NAME:                 $AKS_MGMT_SUBNET_NAME"
+  echo "AKS_MGMT_SUBNET_CIDR:                 $AKS_MGMT_SUBNET_CIDR"
 
   echo "AZURE_SUBSCRIPTION_ID:                $AZURE_SUBSCRIPTION_ID"
   echo "AZURE_CLIENT_ID:                      $AZURE_CLIENT_ID"
@@ -102,6 +115,16 @@ create_aks_cluster() {
     --location "${AZURE_LOCATION}" \
     --output none --only-show-errors \
     --tags creationTimestamp="${TIMESTAMP}" jobName="${JOB_NAME}" buildProvenance="${BUILD_PROVENANCE}"
+
+    echo "creating vnet for the resource group ${AKS_RESOURCE_GROUP}"
+    az network vnet create \
+      --resource-group "${AKS_RESOURCE_GROUP}"\
+      --name "${AKS_MGMT_VNET_NAME}" \
+      --address-prefix "${AKS_MGMT_VNET_CIDR}" \
+      --subnet-name "${AKS_MGMT_SUBNET_NAME}" \
+      --subnet-prefix "${AKS_MGMT_SUBNET_CIDR}" \
+      --output none --only-show-errors \
+      --tags creationTimestamp="${TIMESTAMP}" jobName="${JOB_NAME}" buildProvenance="${BUILD_PROVENANCE}"
   fi
 
   aks_exists=$(az aks show --name "${MGMT_CLUSTER_NAME}" --resource-group "${AKS_RESOURCE_GROUP}" 2>&1 || true) # true because we want to continue if the command fails
@@ -110,13 +133,16 @@ create_aks_cluster() {
     az aks create --name "${MGMT_CLUSTER_NAME}" \
     --resource-group "${AKS_RESOURCE_GROUP}" \
     --location "${AZURE_LOCATION}" \
-    --kubernetes-version "${KUBERNETES_VERSION}" \
+    --kubernetes-version "${AKS_MGMT_KUBERNETES_VERSION}" \
     --node-count "${AKS_NODE_COUNT}" \
     --node-vm-size "${AKS_NODE_VM_SIZE}" \
     --node-resource-group "${AKS_NODE_RESOURCE_GROUP}" \
     --vm-set-type VirtualMachineScaleSets \
     --generate-ssh-keys \
     --network-plugin azure \
+    --vnet-subnet-id "/subscriptions/${AZURE_SUBSCRIPTION_ID}/resourceGroups/${AKS_RESOURCE_GROUP}/providers/Microsoft.Network/virtualNetworks/${AKS_MGMT_VNET_NAME}/subnets/${AKS_MGMT_SUBNET_NAME}" \
+    --service-cidr "${AKS_MGMT_SERVICE_CIDR}" \
+    --dns-service-ip "${AKS_MGMT_DNS_SERVICE_IP}" \
     --tags creationTimestamp="${TIMESTAMP}" jobName="${JOB_NAME}" buildProvenance="${BUILD_PROVENANCE}" \
     --output none --only-show-errors;
   elif echo "$aks_exists" | grep -q "${MGMT_CLUSTER_NAME}"; then
@@ -127,6 +153,7 @@ create_aks_cluster() {
   fi
 
   # check and save kubeconfig
+  echo -e "\n"
   echo "saving credentials of cluster ${MGMT_CLUSTER_NAME} in ${REPO_ROOT}/${MGMT_CLUSTER_KUBECONFIG}"
   az aks get-credentials --name "${MGMT_CLUSTER_NAME}" --resource-group "${AKS_RESOURCE_GROUP}" \
   --file "${REPO_ROOT}/${MGMT_CLUSTER_KUBECONFIG}" --only-show-errors
@@ -179,15 +206,10 @@ create_aks_cluster() {
 set_env_varaibles(){
   cat <<EOF > tilt-settings-temp.yaml
 kustomize_substitutions:
-  MGMT_CLUSTER_NAME: "${MGMT_CLUSTER_NAME}"
   AKS_RESOURCE_GROUP: "${AKS_RESOURCE_GROUP}"
   AKS_NODE_RESOURCE_GROUP: "${AKS_NODE_RESOURCE_GROUP}"
-  MGMT_CLUSTER_KUBECONFIG: "${MGMT_CLUSTER_KUBECONFIG}"
-  AKS_MI_CLIENT_ID: "${AKS_MI_CLIENT_ID}"
-  AKS_MI_OBJECT_ID: "${AKS_MI_OBJECT_ID}"
-  AKS_MI_RESOURCE_ID: "${AKS_MI_RESOURCE_ID}"
-  MANAGED_IDENTITY_NAME: "${MANAGED_IDENTITY_NAME}"
-  MANAGED_IDENTITY_RG: "${MANAGED_IDENTITY_RG}"
+  AKS_MGMT_VNET_NAME: "${AKS_MGMT_VNET_NAME}"
+  MGMT_CLUSTER_NAME: "${MGMT_CLUSTER_NAME}"
   AZURE_CLIENT_ID_USER_ASSIGNED_IDENTITY: "${AKS_MI_CLIENT_ID}"
   CI_RG: "${MANAGED_IDENTITY_RG}"
   USER_IDENTITY: "${MANAGED_IDENTITY_NAME}"

diff --git a/templates/cluster-template-aad.yaml b/templates/cluster-template-aad.yaml
diff --git a/templates/cluster-template-azure-bastion.yaml b/templates/cluster-template-azure-bastion.yaml