Add dev-gke-noauth overlay that allows dev version of driver to be deployed

cmd/gce-pd-csi-driver/main.go

@@ -94,11 +94,9 @@ var (
 
 	extraTagsStr = flag.String("extra-tags", "", "Extra tags to attach to each Compute Disk, Image, Snapshot created. It is a comma separated list of parent id, key and value like '<parent_id1>/<tag_key1>/<tag_value1>,...,<parent_idN>/<tag_keyN>/<tag_valueN>'. parent_id is the Organization or the Project ID or Project name where the tag key and the tag value resources exist. A maximum of 50 tags bindings is allowed for a resource. See https://cloud.google.com/resource-manager/docs/tags/tags-overview, https://cloud.google.com/resource-manager/docs/tags/tags-creating-and-managing for details")
 
-	version string
-)
+	driverName = flag.String("driver-name", "pd.csi.storage.gke.io", "Driver name. Defaults to pd.csi.storage.gke.io")
 
-const (
-	driverName = "pd.csi.storage.gke.io"
+	version string
 )
 
 func init() {
@@ -150,7 +148,7 @@ func handle() {
 
 	var metricsManager *metrics.MetricsManager = nil
 	if *runControllerService && *httpEndpoint != "" {
-		mm := metrics.NewMetricsManager()
+		mm := metrics.NewMetricsManager(*driverName)
 		mm.InitializeHttpHandler(*httpEndpoint, *metricsPath)
 		mm.RegisterPDCSIMetric()
 
@@ -272,7 +270,7 @@ func handle() {
 		}
 	}
 
-	err = gceDriver.SetupGCEDriver(driverName, version, extraVolumeLabels, extraTags, identityServer, controllerServer, nodeServer)
+	err = gceDriver.SetupGCEDriver(*driverName, version, extraVolumeLabels, extraTags, identityServer, controllerServer, nodeServer)
 	if err != nil {
 		klog.Fatalf("Failed to initialize GCE CSI Driver: %v", err.Error())
 	}

deploy/kubernetes/base/controller/controller.yaml

@@ -16,7 +16,7 @@ spec:
       # since it replaces GCE Metadata Server with GKE Metadata Server. Remove
       # this requirement when issue is resolved and before any exposure of
       # metrics ports
-      hostNetwork: true
+      hostNetwork: false
       nodeSelector:
         kubernetes.io/os: linux
       serviceAccountName: csi-gce-pd-controller-sa
@@ -145,6 +145,9 @@ spec:
             - "--supports-dynamic-iops-provisioning=hyperdisk-balanced,hyperdisk-extreme"
             - "--supports-dynamic-throughput-provisioning=hyperdisk-balanced,hyperdisk-throughput,hyperdisk-ml"
             - --enable-data-cache
+            - --run-node-service=false
+            - --multi-zone-volume-handle-disk-types=hyperdisk-ml
+            - --multi-zone-volume-handle-enable
           command:
             - /gce-pd-csi-driver
           env:

deploy/kubernetes/overlays/dev-gke-noauth/controller_always_pull.yaml

@@ -0,0 +1,10 @@
+kind: Deployment
+apiVersion: apps/v1
+metadata:
+  name: csi-gce-pd-controller
+spec:
+  template:
+    spec:
+      containers:
+        - name: gce-pd-driver
+          imagePullPolicy: Always

deploy/kubernetes/overlays/dev-gke-noauth/csidriver_dev_name.yaml

@@ -0,0 +1,3 @@
+- op: replace
+  path: /metadata/name
+  value: pd.dev.csi.storage.gke.io

deploy/kubernetes/overlays/dev-gke-noauth/driver-args.yaml

@@ -0,0 +1,3 @@
+- op: add
+  path: /spec/template/spec/containers/0/args/-
+  value: --driver-name=pd.dev.csi.storage.gke.io

deploy/kubernetes/overlays/dev-gke-noauth/kustomization.yaml

@@ -0,0 +1,36 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+patchesStrategicMerge:
+- controller_always_pull.yaml
+- node_always_pull.yaml
+- noauth.yaml
+- node_overwrite_pdcsi_path.yaml
+namespace: gce-pd-csi-driver
+resources:
+- ../../base/
+# Here dev overlay is using the same image as alpha
+transformers:
+- ../../images/stable-master
+# Apply patches to support dynamic provisioning for hyperdisks
+patches:
+- path: ./driver-args.yaml
+  target:
+    group: apps
+    version: v1
+    kind: Deployment
+    name: csi-gce-pd-controller
+- path: ./node-driver-args.yaml
+  target:
+    group: apps
+    version: v1
+    kind: DaemonSet
+    name: csi-gce-pd-node
+- path: ./csidriver_dev_name.yaml
+  target:
+    kind: CSIDriver
+    name: pd.csi.storage.gke.io
+# To change the dev image, add something like the following.
+images:
+- name: gke.gcr.io/gcp-compute-persistent-disk-csi-driver
+  newName: us-central1-docker.pkg.dev/psch-gke-dev/csi-dev/gcp-compute-persistent-disk-csi-driver
+  newTag: latest

deploy/kubernetes/overlays/dev-gke-noauth/noauth.yaml

@@ -0,0 +1,25 @@
+kind: Deployment
+apiVersion: apps/v1
+metadata:
+  name: csi-gce-pd-controller
+spec:
+  template:
+    spec:
+      containers:
+        - name: gce-pd-driver
+          env:
+            - $patch: delete
+              name: GOOGLE_APPLICATION_CREDENTIALS
+              value: "/etc/cloud-sa/cloud-sa.json"
+          volumeMounts:
+            - $patch: delete
+              name: cloud-sa-volume
+              readOnly: true
+              mountPath: "/etc/cloud-sa"
+      volumes:
+        - $patch: delete
+          name: cloud-sa-volume
+          secret:
+            secretName: cloud-sa
+
+

deploy/kubernetes/overlays/dev-gke-noauth/node-driver-args.yaml

@@ -0,0 +1,3 @@
+- op: add
+  path: /spec/template/spec/containers/1/args/-
+  value: --driver-name=pd.dev.csi.storage.gke.io

deploy/kubernetes/overlays/dev-gke-noauth/node_always_pull.yaml

@@ -0,0 +1,11 @@
+kind: DaemonSet
+apiVersion: apps/v1
+metadata:
+  name: csi-gce-pd-node
+spec:
+  template:
+    spec:
+      containers:
+        - name: gce-pd-driver
+          imagePullPolicy: Always
+

deploy/kubernetes/overlays/dev-gke-noauth/node_overwrite_pdcsi_path.yaml

@@ -0,0 +1,18 @@
+kind: DaemonSet
+apiVersion: apps/v1
+metadata:
+  name: csi-gce-pd-node
+spec:
+  template:
+    spec:
+      containers:
+        - name: csi-driver-registrar
+          args:
+            - "--v=5"
+            - "--csi-address=/csi/csi.sock"
+            - "--kubelet-registration-path=/var/lib/kubelet/plugins/pd.dev.csi.storage.gke.io/csi.sock"
+      volumes:
+        - name: plugin-dir
+          hostPath:
+            path: /var/lib/kubelet/plugins/pd.dev.csi.storage.gke.io/
+            type: DirectoryOrCreate

pkg/metrics/metrics.go

@@ -32,7 +32,6 @@ const (
 	// envGKEPDCSIVersion is an environment variable set in the PDCSI controller manifest
 	// with the current version of the GKE component.
 	envGKEPDCSIVersion               = "GKE_PDCSI_VERSION"
-	pdcsiDriverName                  = "pd.csi.storage.gke.io"
 	DefaultDiskTypeForMetric         = "unknownDiskType"
 	DefaultEnableConfidentialCompute = "unknownConfidentialMode"
 	DefaultEnableStoragePools        = "unknownStoragePools"
@@ -56,12 +55,14 @@ var (
 )
 
 type MetricsManager struct {
-	registry metrics.KubeRegistry
+	registry   metrics.KubeRegistry
+	driverName string
 }
 
-func NewMetricsManager() MetricsManager {
+func NewMetricsManager(driverName string) MetricsManager {
 	mm := MetricsManager{
-		registry: metrics.NewKubeRegistry(),
+		driverName: driverName,
+		registry:   metrics.NewKubeRegistry(),
 	}
 	return mm
 }
@@ -97,7 +98,7 @@ func (mm *MetricsManager) RecordOperationErrorMetrics(
 	enableConfidentialStorage string,
 	enableStoragePools string) {
 	errCode := errorCodeLabelValue(operationErr)
-	pdcsiOperationErrorsMetric.WithLabelValues(pdcsiDriverName, fullMethodName, errCode, diskType, enableConfidentialStorage, enableStoragePools).Inc()
+	pdcsiOperationErrorsMetric.WithLabelValues(mm.driverName, fullMethodName, errCode, diskType, enableConfidentialStorage, enableStoragePools).Inc()
 	klog.Infof("Recorded PDCSI operation error code: %q", errCode)
 }