Merge pull request #334 from jacobwolfaws/release-0.10

Cherry-picking commits from master
kubernetes-sigs · Jun 22, 2023 · 307682f · 307682f
2 parents dc97e0a + 35abe3e
commit 307682f
Show file tree

Hide file tree

Showing 23 changed files with 182 additions and 73 deletions.
diff --git a/.github/ISSUE_TEMPLATE/bug_report.md b/.github/ISSUE_TEMPLATE/bug_report.md
@@ -1,6 +1,6 @@
 ---
 name: Bug report
-about: Create a report to help us improve EBS CSI Driver
+about: Create a report to help us improve the FSx for Lustre CSI Driver
 labels: 
 
 ---

diff --git a/charts/aws-fsx-csi-driver/templates/controller-deployment.yaml b/charts/aws-fsx-csi-driver/templates/controller-deployment.yaml
@@ -31,7 +31,6 @@ spec:
         {{- with .Values.controller.nodeSelector }}
         {{- toYaml . | nindent 8 }}
         {{- end }}
-      hostNetwork: true
       serviceAccountName: {{ .Values.controller.serviceAccount.name }}
       priorityClassName: system-cluster-critical
       tolerations:
@@ -66,6 +65,10 @@ spec:
                   name: aws-secret
                   key: access_key
                   optional: true
+            {{- with .Values.controller.region }}
+            - name: AWS_REGION
+              value: {{ . }}
+            {{- end }}
           volumeMounts:
             - name: socket-dir
               mountPath: /var/lib/csi/sockets/pluginproxy/

diff --git a/charts/aws-fsx-csi-driver/templates/node-daemonset.yaml b/charts/aws-fsx-csi-driver/templates/node-daemonset.yaml
@@ -30,7 +30,6 @@ spec:
         {{- with .Values.node.nodeSelector }}
         {{- toYaml . | nindent 8 }}
         {{- end }}
-      hostNetwork: true
       dnsPolicy: {{ .Values.node.dnsPolicy }}
       {{- with .Values.node.dnsConfig }}
       dnsConfig: {{- toYaml . | nindent 8 }}

diff --git a/charts/aws-fsx-csi-driver/values.yaml b/charts/aws-fsx-csi-driver/values.yaml
@@ -14,27 +14,27 @@ sidecars:
   livenessProbe:
     image:
       repository: public.ecr.aws/eks-distro/kubernetes-csi/livenessprobe
-      tag: v2.10.0-eks-1-26-9
+      tag: v2.10.0-eks-1-27-5
       pullPolicy: IfNotPresent
     resources: {}
   nodeDriverRegistrar:
     image:
       repository: public.ecr.aws/eks-distro/kubernetes-csi/node-driver-registrar
-      tag: v2.8.0-eks-1-26-9
+      tag: v2.8.0-eks-1-27-5
       pullPolicy: IfNotPresent
     logLevel: 2
     resources: {}
   provisioner:
     image:
       repository: public.ecr.aws/eks-distro/kubernetes-csi/external-provisioner
-      tag: v3.5.0-eks-1-26-9
+      tag: v3.5.0-eks-1-27-5
       pullPolicy: IfNotPresent
     logLevel: 2
     resources: {}
   resizer:
     image:
       repository: public.ecr.aws/eks-distro/kubernetes-csi/external-resizer
-      tag: v1.8.0-eks-1-26-9
+      tag: v1.8.0-eks-1-27-5
       pullPolicy: IfNotPresent
     logLevel: 2
     resources: {}
@@ -54,6 +54,11 @@ controller:
     annotations: {}
   logLevel: 2
   podAnnotations: {}
+  # AWS region to use. If not specified then the region will be looked up via the AWS EC2 metadata
+  # service.
+  # ---
+  # region: us-east-1
+  region:
   tolerations:
     - key: CriticalAddonsOnly
       operator: Exists

diff --git a/cmd/main.go b/cmd/main.go
@@ -37,6 +37,7 @@ func main() {
 	drv, err := driver.NewDriver(
 		driver.WithEndpoint(options.ServerOptions.Endpoint),
 		driver.WithMode(options.ServerOptions.DriverMode),
+		driver.WithExtraTags(options.ControllerOptions.ExtraTags),
 	)
 
 	if err != nil {

diff --git a/cmd/options/controller_options.go b/cmd/options/controller_options.go
@@ -22,7 +22,10 @@ import (
 
 // ControllerOptions contains options and configuration settings for the controller service.
 type ControllerOptions struct {
+	// ExtraTags is a map of tags that will be attached to each dynamically provisioned resource.
+	ExtraTags string
 }
 
 func (s *ControllerOptions) AddFlags(fs *flag.FlagSet) {
+	fs.StringVar(&s.ExtraTags, "extra-tags", "", "Extra tags to attach to each dynamically provisioned resource. It is a comma separated list of key value pairs like '<key1>=<value1>,<key2>=<value2>'")
 }
diff --git a/deploy/kubernetes/base/controller-deployment.yaml b/deploy/kubernetes/base/controller-deployment.yaml
@@ -20,7 +20,6 @@ spec:
     spec:
       nodeSelector:
         kubernetes.io/os: linux
-      hostNetwork: true
       serviceAccountName: fsx-csi-controller-sa
       priorityClassName: system-cluster-critical
       tolerations:
@@ -73,7 +72,7 @@ spec:
             periodSeconds: 2
             failureThreshold: 5
         - name: csi-provisioner
-          image: public.ecr.aws/eks-distro/kubernetes-csi/external-provisioner:v3.5.0-eks-1-26-9
+          image: public.ecr.aws/eks-distro/kubernetes-csi/external-provisioner:v3.5.0-eks-1-27-5
           args:
             - --csi-address=$(ADDRESS)
             - --v=2
@@ -87,7 +86,7 @@ spec:
             - name: socket-dir
               mountPath: /var/lib/csi/sockets/pluginproxy/
         - name: csi-resizer
-          image: public.ecr.aws/eks-distro/kubernetes-csi/external-resizer:v1.8.0-eks-1-26-9
+          image: public.ecr.aws/eks-distro/kubernetes-csi/external-resizer:v1.8.0-eks-1-27-5
           args:
             - --csi-address=$(ADDRESS)
             - --v=2
@@ -100,7 +99,7 @@ spec:
             - name: socket-dir
               mountPath: /var/lib/csi/sockets/pluginproxy/
         - name: liveness-probe
-          image: public.ecr.aws/eks-distro/kubernetes-csi/livenessprobe:v2.10.0-eks-1-26-9
+          image: public.ecr.aws/eks-distro/kubernetes-csi/livenessprobe:v2.10.0-eks-1-27-5
           args:
             - --csi-address=/csi/csi.sock
             - --health-port=9910

diff --git a/deploy/kubernetes/base/node-daemonset.yaml b/deploy/kubernetes/base/node-daemonset.yaml
@@ -19,7 +19,6 @@ spec:
     spec:
       nodeSelector:
         kubernetes.io/os: linux
-      hostNetwork: true
       dnsPolicy: ClusterFirst
       serviceAccountName: fsx-csi-node-sa
       priorityClassName: system-node-critical
@@ -62,7 +61,7 @@ spec:
             periodSeconds: 2
             failureThreshold: 5
         - name: node-driver-registrar
-          image: public.ecr.aws/eks-distro/kubernetes-csi/node-driver-registrar:v2.8.0-eks-1-26-9
+          image: public.ecr.aws/eks-distro/kubernetes-csi/node-driver-registrar:v2.8.0-eks-1-27-5
           imagePullPolicy: IfNotPresent
           args:
             - --csi-address=$(ADDRESS)
@@ -83,7 +82,7 @@ spec:
             - name: registration-dir
               mountPath: /registration
         - name: liveness-probe
-          image: public.ecr.aws/eks-distro/kubernetes-csi/livenessprobe:v2.10.0-eks-1-26-9
+          image: public.ecr.aws/eks-distro/kubernetes-csi/livenessprobe:v2.10.0-eks-1-27-5
           imagePullPolicy: IfNotPresent
           args:
             - --csi-address=/csi/csi.sock

diff --git a/docs/options.md b/docs/options.md
@@ -1,7 +1,8 @@
 # Driver Options
 There are a couple of driver options that can be passed as arguments when starting the driver container.
 
-| Option argument             | value sample                                      | default                                             | Description         |
-|-----------------------------|---------------------------------------------------|-----------------------------------------------------|---------------------|
-| endpoint                    | tcp://127.0.0.1:10000/                            | unix:///var/lib/csi/sockets/pluginproxy/csi.sock    | The socket on which the driver will listen for CSI RPCs|
-| logging-format              | json                                              | text                                                | Sets the log format. Permitted formats: text, json|
+| Option argument             | value sample                                      | default                                             | Description                                                                                 |
+|-----------------------------|---------------------------------------------------|-----------------------------------------------------|---------------------------------------------------------------------------------------------|
+| endpoint                    | tcp://127.0.0.1:10000/                            | unix:///var/lib/csi/sockets/pluginproxy/csi.sock    | The socket on which the driver will listen for CSI RPCs                                     |
+| extra-tags                  | key1=value1,key2=value2                           |                                                     | Tags specified in the controller spec are attached to each dynamically provisioned resource |
+| logging-format              | json                                              | text                                                | Sets the log format. Permitted formats: text, json                                          |
diff --git a/examples/kubernetes/dynamic_provisioning/specs/pod.yaml b/examples/kubernetes/dynamic_provisioning/specs/pod.yaml
@@ -5,7 +5,7 @@ metadata:
 spec:
   containers:
   - name: app
-    image: centos
+    image: amazonlinux:2
     command: ["/bin/sh"]
     args: ["-c", "while true; do echo $(date -u) >> /data/out.txt; sleep 5; done"]
     volumeMounts:

diff --git a/examples/kubernetes/dynamic_provisioning_s3/specs/pod.yaml b/examples/kubernetes/dynamic_provisioning_s3/specs/pod.yaml
@@ -13,7 +13,7 @@ spec:
     lifecycle:
       postStart:
         exec:
-          command: ["amazon-linux-extras", "install", "lustre2.10", "-y"]
+          command: ["amazon-linux-extras", "install", "lustre", "-y"]
     volumeMounts:
     - name: persistent-storage
       mountPath: /data

diff --git a/examples/kubernetes/max_cache_tuning/README.md b/examples/kubernetes/max_cache_tuning/README.md
@@ -10,11 +10,11 @@ metadata:
 spec:
   initContainers:
   - name: set-lustre-cache
-    image: public.ecr.aws/fsx-csi-driver/aws-fsx-csi-driver:v0.9.0
+    image: amazonlinux:2
+    command: ["/bin/sh","-c"]
+    args: ["amazon-linux-extras install lustre -y && /sbin/lctl set_param llite.*.max_cached_mb=32"]
     securityContext:
       privileged: true
-    command: ["/sbin/lctl"]
-    args: ["set_param", "llite.*.max_cached_mb=32"]
   containers:
   - name: app
     image: amazonlinux:2

diff --git a/examples/kubernetes/max_cache_tuning/specs/pod.yaml b/examples/kubernetes/max_cache_tuning/specs/pod.yaml
@@ -5,11 +5,11 @@ metadata:
 spec:
   initContainers:
   - name: set-lustre-cache
-    image: public.ecr.aws/fsx-csi-driver/aws-fsx-csi-driver:v0.9.0
+    image: amazonlinux:2
+    command: ["/bin/sh","-c"]
+    args: ["amazon-linux-extras install lustre -y && /sbin/lctl set_param llite.*.max_cached_mb=32"]
     securityContext:
       privileged: true
-    command: ["/sbin/lctl"]
-    args: ["set_param", "llite.*.max_cached_mb=32"]
   containers:
   - name: app
     image: amazonlinux:2

diff --git a/examples/kubernetes/multiple_pods/specs/pod1.yaml b/examples/kubernetes/multiple_pods/specs/pod1.yaml
@@ -5,7 +5,7 @@ metadata:
 spec:
   containers:
   - name: app1
-    image: centos
+    image: amazonlinux:2
     command: ["/bin/sh"]
     args: ["-c", "while true; do echo $(date -u) >> /data/out1.txt; sleep 5; done"]
     volumeMounts:

diff --git a/examples/kubernetes/multiple_pods/specs/pod2.yaml b/examples/kubernetes/multiple_pods/specs/pod2.yaml
@@ -5,7 +5,7 @@ metadata:
 spec:
   containers:
   - name: app2
-    image: busybox
+    image: amazonlinux:2
     command: ["/bin/sh"]
     args: ["-c", "while true; do echo $(date -u) >> /data/out2.txt; sleep 5; done"]
     volumeMounts:

diff --git a/examples/kubernetes/static_provisioning/specs/pod.yaml b/examples/kubernetes/static_provisioning/specs/pod.yaml
@@ -5,7 +5,7 @@ metadata:
 spec:
   containers:
   - name: app
-    image: centos
+    image: amazonlinux:2
     command: ["/bin/sh"]
     args: ["-c", "while true; do echo $(date -u) >> /data/out.txt; sleep 5; done"]
     volumeMounts:

diff --git a/hack/kops-patch-node.yaml b/hack/kops-patch-node.yaml
@@ -1,2 +1,3 @@
 spec:
-  image: 137112412989/amzn2-ami-hvm-2.0.20210525.0-x86_64-gp2
+  image: 137112412989/amzn2-ami-hvm-2.0.20220218.3-x86_64-gp2
+
diff --git a/pkg/cloud/cloud.go b/pkg/cloud/cloud.go
@@ -235,9 +235,6 @@ func (c *cloud) CreateFileSystem(ctx context.Context, volumeName string, fileSys
 
 	output, err := c.fsx.CreateFileSystemWithContext(ctx, input)
 	if err != nil {
-		if isIncompatibleParameter(err) {
-			return nil, ErrFsExistsDiffSize
-		}
 		return nil, fmt.Errorf("CreateFileSystem failed: %v", err)
 	}
 
@@ -338,7 +335,7 @@ func (c *cloud) WaitForFileSystemAvailable(ctx context.Context, fileSystemId str
 		if err != nil {
 			return true, err
 		}
-		klog.V(2).Infof("WaitForFileSystemAvailable filesystem %q status is: %q", fileSystemId, *fs.Lifecycle)
+		klog.V(2).InfoS("WaitForFileSystemAvailable", "filesystem", fileSystemId, "status", *fs.Lifecycle)
 		switch *fs.Lifecycle {
 		case "AVAILABLE":
 			return true, nil
@@ -361,7 +358,7 @@ func (c *cloud) WaitForFileSystemResize(ctx context.Context, fileSystemId string
 			return true, err
 		}
 
-		klog.V(2).Infof("WaitForFileSystemResize filesystem %q update status is: %q", fileSystemId, *updateAction.Status)
+		klog.V(2).InfoS("WaitForFileSystemResize", "filesystem", fileSystemId, "update status", *updateAction.Status)
 		switch *updateAction.Status {
 		case "PENDING", "IN_PROGRESS":
 			// The resizing workflow has not completed
@@ -433,15 +430,6 @@ func isFileSystemNotFound(err error) bool {
 	return false
 }
 
-func isIncompatibleParameter(err error) bool {
-	if awsErr, ok := err.(awserr.Error); ok {
-		if awsErr.Code() == fsx.ErrCodeIncompatibleParameterError {
-			return true
-		}
-	}
-	return false
-}
-
 // isBadRequestUpdateInProgress identifies an error returned from the FSx API as a BadRequest with an "update already
 // in progress" message.
 func isBadRequestUpdateInProgress(err error) bool {

diff --git a/pkg/driver/controller.go b/pkg/driver/controller.go
@@ -106,7 +106,7 @@ func newControllerService(driverOptions *DriverOptions) controllerService {
 	}
 }
 func (d *controllerService) CreateVolume(ctx context.Context, req *csi.CreateVolumeRequest) (*csi.CreateVolumeResponse, error) {
-	klog.V(4).Infof("CreateVolume: called with args %#v", req)
+	klog.V(4).InfoS("CreateVolume: called", "args", *req)
 	volName := req.GetName()
 	if len(volName) == 0 {
 		return nil, status.Error(codes.InvalidArgument, "Volume name not provided")
@@ -213,10 +213,18 @@ func (d *controllerService) CreateVolume(ctx context.Context, req *csi.CreateVol
 		fsOptions.CapacityGiB = util.RoundUpVolumeSize(capRange.GetRequiredBytes(), fsOptions.DeploymentType, fsOptions.StorageType, fsOptions.PerUnitStorageThroughput)
 	}
 
+	var tagArray []string
+	optionsTags := d.driverOptions.extraTags
+
+	if optionsTags != "" {
+		tagArray = strings.Split(optionsTags, ",")
+	}
+
 	if val, ok := volumeParams[volumeParamsExtraTags]; ok {
 		extraTags := strings.Split(val, ",")
-		fsOptions.ExtraTags = extraTags
+		tagArray = append(tagArray, extraTags...)
 	}
+	fsOptions.ExtraTags = tagArray
 
 	fs, err := d.cloud.CreateFileSystem(ctx, volName, fsOptions)
 	if err != nil {
@@ -237,7 +245,7 @@ func (d *controllerService) CreateVolume(ctx context.Context, req *csi.CreateVol
 }
 
 func (d *controllerService) DeleteVolume(ctx context.Context, req *csi.DeleteVolumeRequest) (*csi.DeleteVolumeResponse, error) {
-	klog.V(4).Infof("DeleteVolume: called with args: %#v", req)
+	klog.V(4).InfoS("DeleteVolume: called", "args", *req)
 	volumeID := req.GetVolumeId()
 	if len(volumeID) == 0 {
 		return nil, status.Error(codes.InvalidArgument, "Volume ID not provided")
@@ -252,7 +260,7 @@ func (d *controllerService) DeleteVolume(ctx context.Context, req *csi.DeleteVol
 
 	if err := d.cloud.DeleteFileSystem(ctx, volumeID); err != nil {
 		if err == cloud.ErrNotFound {
-			klog.V(4).Infof("DeleteVolume: volume not found, returning with success")
+			klog.V(4).InfoS("DeleteVolume: volume not found, returning with success")
 			return &csi.DeleteVolumeResponse{}, nil
 		}
 		return nil, status.Errorf(codes.Internal, "Could not delete volume ID %q: %v", volumeID, err)
@@ -269,7 +277,7 @@ func (d *controllerService) ControllerUnpublishVolume(ctx context.Context, req *
 }
 
 func (d *controllerService) ControllerGetCapabilities(ctx context.Context, req *csi.ControllerGetCapabilitiesRequest) (*csi.ControllerGetCapabilitiesResponse, error) {
-	klog.V(4).Infof("ControllerGetCapabilities: called with args %#v", req)
+	klog.V(4).InfoS("ControllerGetCapabilities: called", "args", *req)
 	var caps []*csi.ControllerServiceCapability
 	for _, cap := range controllerCaps {
 		c := &csi.ControllerServiceCapability{
@@ -285,17 +293,17 @@ func (d *controllerService) ControllerGetCapabilities(ctx context.Context, req *
 }
 
 func (d *controllerService) GetCapacity(ctx context.Context, req *csi.GetCapacityRequest) (*csi.GetCapacityResponse, error) {
-	klog.V(4).Infof("GetCapacity: called with args %#v", req)
+	klog.V(4).InfoS("GetCapacity: called", "args", *req)
 	return nil, status.Error(codes.Unimplemented, "")
 }
 
 func (d *controllerService) ListVolumes(ctx context.Context, req *csi.ListVolumesRequest) (*csi.ListVolumesResponse, error) {
-	klog.V(4).Infof("ListVolumes: called with args %#v", req)
+	klog.V(4).InfoS("ListVolumes: called", "args", *req)
 	return nil, status.Error(codes.Unimplemented, "")
 }
 
 func (d *controllerService) ValidateVolumeCapabilities(ctx context.Context, req *csi.ValidateVolumeCapabilitiesRequest) (*csi.ValidateVolumeCapabilitiesResponse, error) {
-	klog.V(4).Infof("ValidateVolumeCapabilities: called with args %#v", req)
+	klog.V(4).InfoS("ValidateVolumeCapabilities: called", "args", *req)
 	volumeID := req.GetVolumeId()
 	if len(volumeID) == 0 {
 		return nil, status.Error(codes.InvalidArgument, "Volume ID not provided")
@@ -361,7 +369,7 @@ func (d *controllerService) ListSnapshots(ctx context.Context, req *csi.ListSnap
 }
 
 func (d *controllerService) ControllerExpandVolume(ctx context.Context, req *csi.ControllerExpandVolumeRequest) (*csi.ControllerExpandVolumeResponse, error) {
-	klog.V(4).Infof("ControllerExpandVolume: called with args %+v", *req)
+	klog.V(4).InfoS("ControllerExpandVolume: called", "args", *req)
 	volumeID := req.GetVolumeId()
 	if len(volumeID) == 0 {
 		return nil, status.Error(codes.InvalidArgument, "Volume ID not provided")
@@ -389,7 +397,7 @@ func (d *controllerService) ControllerExpandVolume(ctx context.Context, req *csi
 	}
 	if newSizeGiB <= fs.CapacityGiB {
 		// Current capacity is sufficient to satisfy the request
-		klog.V(4).Infof("ControllerExpandVolume: current filesystem capacity of %d GiB matches or exceeds requested storage capacity of %d GiB, returning with success", fs.CapacityGiB, newSizeGiB)
+		klog.V(4).InfoS("ControllerExpandVolume: current filesystem capacity matches or exceeds requested storage capacity, returning with success", "current capacity", fs.CapacityGiB, "requested capacity", newSizeGiB)
 		return &csi.ControllerExpandVolumeResponse{
 			CapacityBytes:         util.GiBToBytes(fs.CapacityGiB),
 			NodeExpansionRequired: false,