diff --git a/pkg/kubelet/kubelet.go b/pkg/kubelet/kubelet.go
index 0afc400bc3d..3d89c070848 100644
--- a/pkg/kubelet/kubelet.go
+++ b/pkg/kubelet/kubelet.go
@@ -1983,7 +1983,7 @@ func (kl *Kubelet) cleanupOrphanedVolumes(pods []*api.Pod, runningPods []*kubeco
 		runningSet.Insert(string(pod.ID))
 	}
 
-	for name, cleanerTuple := range currentVolumes {
+	for name, cleaner := range currentVolumes {
 		if _, ok := desiredVolumes[name]; !ok {
 			parts := strings.Split(name, "/")
 			if runningSet.Has(parts[0]) {
@@ -1996,19 +1996,19 @@ func (kl *Kubelet) cleanupOrphanedVolumes(pods []*api.Pod, runningPods []*kubeco
 			// TODO(yifan): Refactor this hacky string manipulation.
 			kl.volumeManager.DeleteVolumes(types.UID(parts[0]))
 			// Get path reference count
-			refs, err := mount.GetMountRefs(kl.mounter, cleanerTuple.Unmounter.GetPath())
+			refs, err := mount.GetMountRefs(kl.mounter, cleaner.Unmounter.GetPath())
 			if err != nil {
 				return fmt.Errorf("Could not get mount path references %v", err)
 			}
 			//TODO (jonesdl) This should not block other kubelet synchronization procedures
-			err = cleanerTuple.Unmounter.TearDown()
+			err = cleaner.Unmounter.TearDown()
 			if err != nil {
 				glog.Errorf("Could not tear down volume %q: %v", name, err)
 			}
 
 			// volume is unmounted.  some volumes also require detachment from the node.
-			if cleanerTuple.Detacher != nil && len(refs) == 1 {
-				detacher := *cleanerTuple.Detacher
+			if cleaner.Detacher != nil && len(refs) == 1 {
+				detacher := *cleaner.Detacher
 				err = detacher.Detach()
 				if err != nil {
 					glog.Errorf("Could not detach volume %q: %v", name, err)
diff --git a/pkg/kubelet/volumes.go b/pkg/kubelet/volumes.go
index 968913389c9..e8c2f38d08a 100644
--- a/pkg/kubelet/volumes.go
+++ b/pkg/kubelet/volumes.go
@@ -207,9 +207,9 @@ func (kl *Kubelet) getPodVolumes(podUID types.UID) ([]*volumeTuple, error) {
 	return volumes, nil
 }
 
-// cleanerTuple is a union struct to allow separating detaching from the cleaner.
+// cleaner is a union struct to allow separating detaching from the cleaner.
 // some volumes require detachment but not all.  Unmounter cannot be nil but Detacher is optional.
-type cleanerTuple struct {
+type cleaner struct {
 	Unmounter volume.Unmounter
 	Detacher  *volume.Detacher
 }
@@ -217,12 +217,12 @@ type cleanerTuple struct {
 // getPodVolumesFromDisk examines directory structure to determine volumes that
 // are presently active and mounted. Returns a union struct containing a volume.Unmounter
 // and potentially a volume.Detacher.
-func (kl *Kubelet) getPodVolumesFromDisk() map[string]cleanerTuple {
-	currentVolumes := make(map[string]cleanerTuple)
+func (kl *Kubelet) getPodVolumesFromDisk() map[string]cleaner {
+	currentVolumes := make(map[string]cleaner)
 	podUIDs, err := kl.listPodsFromDisk()
 	if err != nil {
 		glog.Errorf("Could not get pods from disk: %v", err)
-		return map[string]cleanerTuple{}
+		return map[string]cleaner{}
 	}
 	// Find the volumes for each on-disk pod.
 	for _, podUID := range podUIDs {
@@ -245,7 +245,7 @@ func (kl *Kubelet) getPodVolumesFromDisk() map[string]cleanerTuple {
 				continue
 			}
 
-			tuple := cleanerTuple{Unmounter: unmounter}
+			tuple := cleaner{Unmounter: unmounter}
 			detacher, err := kl.newVolumeDetacherFromPlugins(volume.Kind, volume.Name, podUID)
 			// plugin can be nil but a non-nil error is a legitimate error
 			if err != nil {