From 25107c2e114be80740b7cd2af14aab3e2d9dfba2 Mon Sep 17 00:00:00 2001
From: poornas <poornas@users.noreply.github.com>
Date: Tue, 20 Feb 2018 12:21:12 -0800
Subject: [PATCH] Add NAS gateway support (#5516)

---
 .gitignore                         |   1 +
 cmd/bucket-policy.go               |   6 +-
 cmd/fs-v1-metadata_test.go         |   4 +-
 cmd/fs-v1-multipart.go             |  26 +++----
 cmd/fs-v1-multipart_test.go        |  14 ++--
 cmd/fs-v1.go                       |  90 +++++++++++-----------
 cmd/fs-v1_test.go                  |  20 ++---
 cmd/gateway/gateway.go             |   1 +
 cmd/gateway/nas/gateway-nas.go     | 119 +++++++++++++++++++++++++++++
 cmd/object-api-listobjects_test.go |   2 +-
 cmd/server-main.go                 |   2 +-
 cmd/server-main_test.go            |   2 +-
 cmd/test-utils_test.go             |   6 +-
 cmd/xl-sets.go                     |   4 +-
 cmd/xl-v1-bucket.go                |   4 +-
 docs/gateway/azure.md              |   2 +-
 docs/gateway/b2.md                 |   2 +-
 docs/gateway/gcs.md                |   2 +-
 docs/gateway/manta.md              |   2 +-
 docs/gateway/nas.md                |  43 +++++++++++
 docs/gateway/oss.md                |   2 +-
 21 files changed, 259 insertions(+), 95 deletions(-)
 create mode 100644 cmd/gateway/nas/gateway-nas.go
 create mode 100644 docs/gateway/nas.md

diff --git a/.gitignore b/.gitignore
index ca1800fd8..822ee70fe 100644
--- a/.gitignore
+++ b/.gitignore
@@ -24,3 +24,4 @@ prime/
 snap/.snapcraft/
 stage/
 .sia_temp/
+buildcoveragecoverage.txt
\ No newline at end of file
diff --git a/cmd/bucket-policy.go b/cmd/bucket-policy.go
index a225f9338..eaddd5c28 100644
--- a/cmd/bucket-policy.go
+++ b/cmd/bucket-policy.go
@@ -90,7 +90,7 @@ func initBucketPolicies(objAPI ObjectLayer) (*bucketPolicies, error) {
 	policies := make(map[string]policy.BucketAccessPolicy)
 	// Loads bucket policy.
 	for _, bucket := range buckets {
-		bp, pErr := readBucketPolicy(bucket.Name, objAPI)
+		bp, pErr := ReadBucketPolicy(bucket.Name, objAPI)
 		if pErr != nil {
 			// net.Dial fails for rpc client or any
 			// other unexpected errors during net.Dial.
@@ -130,9 +130,9 @@ func readBucketPolicyJSON(bucket string, objAPI ObjectLayer) (bucketPolicyReader
 	return &buffer, nil
 }
 
-// readBucketPolicy - reads bucket policy for an input bucket, returns BucketPolicyNotFound
+// ReadBucketPolicy - reads bucket policy for an input bucket, returns BucketPolicyNotFound
 // if bucket policy is not found. This function also parses the bucket policy into an object.
-func readBucketPolicy(bucket string, objAPI ObjectLayer) (policy.BucketAccessPolicy, error) {
+func ReadBucketPolicy(bucket string, objAPI ObjectLayer) (policy.BucketAccessPolicy, error) {
 	// Read bucket policy JSON.
 	bucketPolicyReader, err := readBucketPolicyJSON(bucket, objAPI)
 	if err != nil {
diff --git a/cmd/fs-v1-metadata_test.go b/cmd/fs-v1-metadata_test.go
index 27773cdf5..22fb9e441 100644
--- a/cmd/fs-v1-metadata_test.go
+++ b/cmd/fs-v1-metadata_test.go
@@ -44,7 +44,7 @@ func TestReadFSMetadata(t *testing.T) {
 	defer os.RemoveAll(disk)
 
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 
 	bucketName := "bucket"
 	objectName := "object"
@@ -79,7 +79,7 @@ func TestWriteFSMetadata(t *testing.T) {
 	defer os.RemoveAll(disk)
 
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 
 	bucketName := "bucket"
 	objectName := "object"
diff --git a/cmd/fs-v1-multipart.go b/cmd/fs-v1-multipart.go
index 79802993e..b6af7efdd 100644
--- a/cmd/fs-v1-multipart.go
+++ b/cmd/fs-v1-multipart.go
@@ -36,22 +36,22 @@ import (
 )
 
 // Returns EXPORT/.minio.sys/multipart/SHA256/UPLOADID
-func (fs *fsObjects) getUploadIDDir(bucket, object, uploadID string) string {
+func (fs *FSObjects) getUploadIDDir(bucket, object, uploadID string) string {
 	return pathJoin(fs.fsPath, minioMetaMultipartBucket, getSHA256Hash([]byte(pathJoin(bucket, object))), uploadID)
 }
 
 // Returns EXPORT/.minio.sys/multipart/SHA256
-func (fs *fsObjects) getMultipartSHADir(bucket, object string) string {
+func (fs *FSObjects) getMultipartSHADir(bucket, object string) string {
 	return pathJoin(fs.fsPath, minioMetaMultipartBucket, getSHA256Hash([]byte(pathJoin(bucket, object))))
 }
 
 // Returns partNumber.etag
-func (fs *fsObjects) encodePartFile(partNumber int, etag string) string {
+func (fs *FSObjects) encodePartFile(partNumber int, etag string) string {
 	return fmt.Sprintf("%.5d.%s", partNumber, etag)
 }
 
 // Returns partNumber and etag
-func (fs *fsObjects) decodePartFile(name string) (partNumber int, etag string, err error) {
+func (fs *FSObjects) decodePartFile(name string) (partNumber int, etag string, err error) {
 	result := strings.Split(name, ".")
 	if len(result) != 2 {
 		return 0, "", errUnexpected
@@ -64,7 +64,7 @@ func (fs *fsObjects) decodePartFile(name string) (partNumber int, etag string, e
 }
 
 // Appends parts to an appendFile sequentially.
-func (fs *fsObjects) backgroundAppend(bucket, object, uploadID string) {
+func (fs *FSObjects) backgroundAppend(bucket, object, uploadID string) {
 	fs.appendFileMapMu.Lock()
 	file := fs.appendFileMap[uploadID]
 	if file == nil {
@@ -121,7 +121,7 @@ func (fs *fsObjects) backgroundAppend(bucket, object, uploadID string) {
 
 // ListMultipartUploads - lists all the uploadIDs for the specified object.
 // We do not support prefix based listing.
-func (fs *fsObjects) ListMultipartUploads(bucket, object, keyMarker, uploadIDMarker, delimiter string, maxUploads int) (result ListMultipartsInfo, e error) {
+func (fs *FSObjects) ListMultipartUploads(bucket, object, keyMarker, uploadIDMarker, delimiter string, maxUploads int) (result ListMultipartsInfo, e error) {
 	if err := checkListMultipartArgs(bucket, object, keyMarker, uploadIDMarker, delimiter, fs); err != nil {
 		return result, toObjectErr(errors.Trace(err))
 	}
@@ -203,7 +203,7 @@ func (fs *fsObjects) ListMultipartUploads(bucket, object, keyMarker, uploadIDMar
 // subsequent request each UUID is unique.
 //
 // Implements S3 compatible initiate multipart API.
-func (fs *fsObjects) NewMultipartUpload(bucket, object string, meta map[string]string) (string, error) {
+func (fs *FSObjects) NewMultipartUpload(bucket, object string, meta map[string]string) (string, error) {
 	if err := checkNewMultipartArgs(bucket, object, fs); err != nil {
 		return "", toObjectErr(err, bucket)
 	}
@@ -238,7 +238,7 @@ func (fs *fsObjects) NewMultipartUpload(bucket, object string, meta map[string]s
 // CopyObjectPart - similar to PutObjectPart but reads data from an existing
 // object. Internally incoming data is written to '.minio.sys/tmp' location
 // and safely renamed to '.minio.sys/multipart' for reach parts.
-func (fs *fsObjects) CopyObjectPart(srcBucket, srcObject, dstBucket, dstObject, uploadID string, partID int,
+func (fs *FSObjects) CopyObjectPart(srcBucket, srcObject, dstBucket, dstObject, uploadID string, partID int,
 	startOffset int64, length int64, metadata map[string]string, srcEtag string) (pi PartInfo, e error) {
 
 	if err := checkNewMultipartArgs(srcBucket, srcObject, fs); err != nil {
@@ -277,7 +277,7 @@ func (fs *fsObjects) CopyObjectPart(srcBucket, srcObject, dstBucket, dstObject,
 // an ongoing multipart transaction. Internally incoming data is
 // written to '.minio.sys/tmp' location and safely renamed to
 // '.minio.sys/multipart' for reach parts.
-func (fs *fsObjects) PutObjectPart(bucket, object, uploadID string, partID int, data *hash.Reader) (pi PartInfo, e error) {
+func (fs *FSObjects) PutObjectPart(bucket, object, uploadID string, partID int, data *hash.Reader) (pi PartInfo, e error) {
 	if err := checkPutObjectPartArgs(bucket, object, fs); err != nil {
 		return pi, toObjectErr(errors.Trace(err), bucket)
 	}
@@ -358,7 +358,7 @@ func (fs *fsObjects) PutObjectPart(bucket, object, uploadID string, partID int,
 // Implements S3 compatible ListObjectParts API. The resulting
 // ListPartsInfo structure is unmarshalled directly into XML and
 // replied back to the client.
-func (fs *fsObjects) ListObjectParts(bucket, object, uploadID string, partNumberMarker, maxParts int) (result ListPartsInfo, e error) {
+func (fs *FSObjects) ListObjectParts(bucket, object, uploadID string, partNumberMarker, maxParts int) (result ListPartsInfo, e error) {
 	if err := checkListPartsArgs(bucket, object, fs); err != nil {
 		return result, toObjectErr(errors.Trace(err))
 	}
@@ -460,7 +460,7 @@ func (fs *fsObjects) ListObjectParts(bucket, object, uploadID string, partNumber
 // md5sums of all the parts.
 //
 // Implements S3 compatible Complete multipart API.
-func (fs *fsObjects) CompleteMultipartUpload(bucket string, object string, uploadID string, parts []CompletePart) (oi ObjectInfo, e error) {
+func (fs *FSObjects) CompleteMultipartUpload(bucket string, object string, uploadID string, parts []CompletePart) (oi ObjectInfo, e error) {
 	if err := checkCompleteMultipartArgs(bucket, object, fs); err != nil {
 		return oi, toObjectErr(err)
 	}
@@ -634,7 +634,7 @@ func (fs *fsObjects) CompleteMultipartUpload(bucket string, object string, uploa
 // that this is an atomic idempotent operation. Subsequent calls have
 // no affect and further requests to the same uploadID would not be
 // honored.
-func (fs *fsObjects) AbortMultipartUpload(bucket, object, uploadID string) error {
+func (fs *FSObjects) AbortMultipartUpload(bucket, object, uploadID string) error {
 	if err := checkAbortMultipartArgs(bucket, object, fs); err != nil {
 		return err
 	}
@@ -666,7 +666,7 @@ func (fs *fsObjects) AbortMultipartUpload(bucket, object, uploadID string) error
 // Removes multipart uploads if any older than `expiry` duration
 // on all buckets for every `cleanupInterval`, this function is
 // blocking and should be run in a go-routine.
-func (fs *fsObjects) cleanupStaleMultipartUploads(cleanupInterval, expiry time.Duration, doneCh chan struct{}) {
+func (fs *FSObjects) cleanupStaleMultipartUploads(cleanupInterval, expiry time.Duration, doneCh chan struct{}) {
 	ticker := time.NewTicker(cleanupInterval)
 	for {
 		select {
diff --git a/cmd/fs-v1-multipart_test.go b/cmd/fs-v1-multipart_test.go
index d56b5331b..04544b21f 100644
--- a/cmd/fs-v1-multipart_test.go
+++ b/cmd/fs-v1-multipart_test.go
@@ -33,7 +33,7 @@ func TestFSCleanupMultipartUploadsInRoutine(t *testing.T) {
 	defer os.RemoveAll(disk)
 
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 
 	// Close the go-routine, we are going to
 	// manually start it and test in this test case.
@@ -73,7 +73,7 @@ func TestNewMultipartUploadFaultyDisk(t *testing.T) {
 	defer os.RemoveAll(disk)
 	obj := initFSObjects(disk, t)
 
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 	objectName := "object"
 
@@ -102,7 +102,7 @@ func TestPutObjectPartFaultyDisk(t *testing.T) {
 	disk := filepath.Join(globalTestTmpDir, "minio-"+nextSuffix())
 	defer os.RemoveAll(disk)
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 	objectName := "object"
 	data := []byte("12345")
@@ -134,7 +134,7 @@ func TestCompleteMultipartUploadFaultyDisk(t *testing.T) {
 	defer os.RemoveAll(disk)
 	obj := initFSObjects(disk, t)
 
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 	objectName := "object"
 	data := []byte("12345")
@@ -166,7 +166,7 @@ func TestCompleteMultipartUpload(t *testing.T) {
 	defer os.RemoveAll(disk)
 	obj := initFSObjects(disk, t)
 
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 	objectName := "object"
 	data := []byte("12345")
@@ -200,7 +200,7 @@ func TestAbortMultipartUpload(t *testing.T) {
 	defer os.RemoveAll(disk)
 	obj := initFSObjects(disk, t)
 
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 	objectName := "object"
 	data := []byte("12345")
@@ -233,7 +233,7 @@ func TestListMultipartUploadsFaultyDisk(t *testing.T) {
 
 	obj := initFSObjects(disk, t)
 
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 	objectName := "object"
 
diff --git a/cmd/fs-v1.go b/cmd/fs-v1.go
index f918e33f2..a2228bcd9 100644
--- a/cmd/fs-v1.go
+++ b/cmd/fs-v1.go
@@ -37,8 +37,8 @@ import (
 	"github.com/minio/minio/pkg/madmin"
 )
 
-// fsObjects - Implements fs object layer.
-type fsObjects struct {
+// FSObjects - Implements fs object layer.
+type FSObjects struct {
 	// Path to be exported over S3 API.
 	fsPath string
 
@@ -93,8 +93,8 @@ func initMetaVolumeFS(fsPath, fsUUID string) error {
 
 }
 
-// newFSObjectLayer - initialize new fs object layer.
-func newFSObjectLayer(fsPath string) (ObjectLayer, error) {
+// NewFSObjectLayer - initialize new fs object layer.
+func NewFSObjectLayer(fsPath string) (ObjectLayer, error) {
 	if fsPath == "" {
 		return nil, errInvalidArgument
 	}
@@ -146,7 +146,7 @@ func newFSObjectLayer(fsPath string) (ObjectLayer, error) {
 	}
 
 	// Initialize fs objects.
-	fs := &fsObjects{
+	fs := &FSObjects{
 		fsPath: fsPath,
 		fsUUID: fsUUID,
 		rwPool: &fsIOPool{
@@ -180,8 +180,8 @@ func newFSObjectLayer(fsPath string) (ObjectLayer, error) {
 	return fs, nil
 }
 
-// Should be called when process shuts down.
-func (fs *fsObjects) Shutdown() error {
+// Shutdown - should be called when process shuts down.
+func (fs *FSObjects) Shutdown() error {
 	fs.fsFormatRlk.Close()
 
 	// Cleanup and delete tmp uuid.
@@ -189,7 +189,7 @@ func (fs *fsObjects) Shutdown() error {
 }
 
 // StorageInfo - returns underlying storage statistics.
-func (fs *fsObjects) StorageInfo() StorageInfo {
+func (fs *FSObjects) StorageInfo() StorageInfo {
 	info, err := getDiskInfo((fs.fsPath))
 	errorIf(err, "Unable to get disk info %#v", fs.fsPath)
 	storageInfo := StorageInfo{
@@ -202,13 +202,13 @@ func (fs *fsObjects) StorageInfo() StorageInfo {
 
 // Locking operations
 
-// List namespace locks held in object layer
-func (fs *fsObjects) ListLocks(bucket, prefix string, duration time.Duration) ([]VolumeLockInfo, error) {
+// ListLocks - List namespace locks held in object layer
+func (fs *FSObjects) ListLocks(bucket, prefix string, duration time.Duration) ([]VolumeLockInfo, error) {
 	return []VolumeLockInfo{}, NotImplemented{}
 }
 
-// Clear namespace locks held in object layer
-func (fs *fsObjects) ClearLocks([]VolumeLockInfo) error {
+// ClearLocks - Clear namespace locks held in object layer
+func (fs *FSObjects) ClearLocks([]VolumeLockInfo) error {
 	return NotImplemented{}
 }
 
@@ -217,7 +217,7 @@ func (fs *fsObjects) ClearLocks([]VolumeLockInfo) error {
 // getBucketDir - will convert incoming bucket names to
 // corresponding valid bucket names on the backend in a platform
 // compatible way for all operating systems.
-func (fs *fsObjects) getBucketDir(bucket string) (string, error) {
+func (fs *FSObjects) getBucketDir(bucket string) (string, error) {
 	// Verify if bucket is valid.
 	if !IsValidBucketName(bucket) {
 		return "", errors.Trace(BucketNameInvalid{Bucket: bucket})
@@ -227,7 +227,7 @@ func (fs *fsObjects) getBucketDir(bucket string) (string, error) {
 	return bucketDir, nil
 }
 
-func (fs *fsObjects) statBucketDir(bucket string) (os.FileInfo, error) {
+func (fs *FSObjects) statBucketDir(bucket string) (os.FileInfo, error) {
 	bucketDir, err := fs.getBucketDir(bucket)
 	if err != nil {
 		return nil, err
@@ -239,9 +239,9 @@ func (fs *fsObjects) statBucketDir(bucket string) (os.FileInfo, error) {
 	return st, nil
 }
 
-// MakeBucket - create a new bucket, returns if it
+// MakeBucketWithLocation - create a new bucket, returns if it
 // already exists.
-func (fs *fsObjects) MakeBucketWithLocation(bucket, location string) error {
+func (fs *FSObjects) MakeBucketWithLocation(bucket, location string) error {
 	bucketLock := fs.nsMutex.NewNSLock(bucket, "")
 	if err := bucketLock.GetLock(globalObjectTimeout); err != nil {
 		return err
@@ -260,7 +260,7 @@ func (fs *fsObjects) MakeBucketWithLocation(bucket, location string) error {
 }
 
 // GetBucketInfo - fetch bucket metadata info.
-func (fs *fsObjects) GetBucketInfo(bucket string) (bi BucketInfo, e error) {
+func (fs *FSObjects) GetBucketInfo(bucket string) (bi BucketInfo, e error) {
 	bucketLock := fs.nsMutex.NewNSLock(bucket, "")
 	if e := bucketLock.GetRLock(globalObjectTimeout); e != nil {
 		return bi, e
@@ -280,7 +280,7 @@ func (fs *fsObjects) GetBucketInfo(bucket string) (bi BucketInfo, e error) {
 }
 
 // ListBuckets - list all s3 compatible buckets (directories) at fsPath.
-func (fs *fsObjects) ListBuckets() ([]BucketInfo, error) {
+func (fs *FSObjects) ListBuckets() ([]BucketInfo, error) {
 	if err := checkPathLength(fs.fsPath); err != nil {
 		return nil, errors.Trace(err)
 	}
@@ -321,7 +321,7 @@ func (fs *fsObjects) ListBuckets() ([]BucketInfo, error) {
 
 // DeleteBucket - delete a bucket and all the metadata associated
 // with the bucket including pending multipart, object metadata.
-func (fs *fsObjects) DeleteBucket(bucket string) error {
+func (fs *FSObjects) DeleteBucket(bucket string) error {
 	bucketLock := fs.nsMutex.NewNSLock(bucket, "")
 	if err := bucketLock.GetLock(globalObjectTimeout); err != nil {
 		return err
@@ -360,7 +360,7 @@ func (fs *fsObjects) DeleteBucket(bucket string) error {
 // CopyObject - copy object source object to destination object.
 // if source object and destination object are same we only
 // update metadata.
-func (fs *fsObjects) CopyObject(srcBucket, srcObject, dstBucket, dstObject string, metadata map[string]string, srcEtag string) (oi ObjectInfo, e error) {
+func (fs *FSObjects) CopyObject(srcBucket, srcObject, dstBucket, dstObject string, metadata map[string]string, srcEtag string) (oi ObjectInfo, e error) {
 	cpSrcDstSame := srcBucket == dstBucket && srcObject == dstObject
 	// Hold write lock on destination since in both cases
 	// - if source and destination are same
@@ -463,7 +463,7 @@ func (fs *fsObjects) CopyObject(srcBucket, srcObject, dstBucket, dstObject strin
 //
 // startOffset indicates the starting read location of the object.
 // length indicates the total length of the object.
-func (fs *fsObjects) GetObject(bucket, object string, offset int64, length int64, writer io.Writer, etag string) (err error) {
+func (fs *FSObjects) GetObject(bucket, object string, offset int64, length int64, writer io.Writer, etag string) (err error) {
 	if err = checkGetObjArgs(bucket, object); err != nil {
 		return err
 	}
@@ -478,7 +478,7 @@ func (fs *fsObjects) GetObject(bucket, object string, offset int64, length int64
 }
 
 // getObject - wrapper for GetObject
-func (fs *fsObjects) getObject(bucket, object string, offset int64, length int64, writer io.Writer, etag string) (err error) {
+func (fs *FSObjects) getObject(bucket, object string, offset int64, length int64, writer io.Writer, etag string) (err error) {
 	if _, err = fs.statBucketDir(bucket); err != nil {
 		return toObjectErr(err, bucket)
 	}
@@ -549,7 +549,7 @@ func (fs *fsObjects) getObject(bucket, object string, offset int64, length int64
 }
 
 // getObjectInfo - wrapper for reading object metadata and constructs ObjectInfo.
-func (fs *fsObjects) getObjectInfo(bucket, object string) (oi ObjectInfo, e error) {
+func (fs *FSObjects) getObjectInfo(bucket, object string) (oi ObjectInfo, e error) {
 	fsMeta := fsMetaV1{}
 	fi, err := fsStatDir(pathJoin(fs.fsPath, bucket, object))
 	if err != nil && errors.Cause(err) != errFileAccessDenied {
@@ -597,7 +597,7 @@ func (fs *fsObjects) getObjectInfo(bucket, object string) (oi ObjectInfo, e erro
 }
 
 // GetObjectInfo - reads object metadata and replies back ObjectInfo.
-func (fs *fsObjects) GetObjectInfo(bucket, object string) (oi ObjectInfo, e error) {
+func (fs *FSObjects) GetObjectInfo(bucket, object string) (oi ObjectInfo, e error) {
 	// Lock the object before reading.
 	objectLock := fs.nsMutex.NewNSLock(bucket, object)
 	if err := objectLock.GetRLock(globalObjectTimeout); err != nil {
@@ -619,7 +619,7 @@ func (fs *fsObjects) GetObjectInfo(bucket, object string) (oi ObjectInfo, e erro
 // This function does the following check, suppose
 // object is "a/b/c/d", stat makes sure that objects ""a/b/c""
 // "a/b" and "a" do not exist.
-func (fs *fsObjects) parentDirIsObject(bucket, parent string) bool {
+func (fs *FSObjects) parentDirIsObject(bucket, parent string) bool {
 	var isParentDirObject func(string) bool
 	isParentDirObject = func(p string) bool {
 		if p == "." || p == "/" {
@@ -640,7 +640,7 @@ func (fs *fsObjects) parentDirIsObject(bucket, parent string) bool {
 // until EOF, writes data directly to configured filesystem path.
 // Additionally writes `fs.json` which carries the necessary metadata
 // for future object operations.
-func (fs *fsObjects) PutObject(bucket string, object string, data *hash.Reader, metadata map[string]string) (objInfo ObjectInfo, retErr error) {
+func (fs *FSObjects) PutObject(bucket string, object string, data *hash.Reader, metadata map[string]string) (objInfo ObjectInfo, retErr error) {
 	if err := checkPutObjectArgs(bucket, object, fs, data.Size()); err != nil {
 		return ObjectInfo{}, err
 	}
@@ -654,7 +654,7 @@ func (fs *fsObjects) PutObject(bucket string, object string, data *hash.Reader,
 }
 
 // putObject - wrapper for PutObject
-func (fs *fsObjects) putObject(bucket string, object string, data *hash.Reader, metadata map[string]string) (objInfo ObjectInfo, retErr error) {
+func (fs *FSObjects) putObject(bucket string, object string, data *hash.Reader, metadata map[string]string) (objInfo ObjectInfo, retErr error) {
 	// No metadata is set, allocate a new one.
 	if metadata == nil {
 		metadata = make(map[string]string)
@@ -778,7 +778,7 @@ func (fs *fsObjects) putObject(bucket string, object string, data *hash.Reader,
 
 // DeleteObject - deletes an object from a bucket, this operation is destructive
 // and there are no rollbacks supported.
-func (fs *fsObjects) DeleteObject(bucket, object string) error {
+func (fs *FSObjects) DeleteObject(bucket, object string) error {
 	// Acquire a write lock before deleting the object.
 	objectLock := fs.nsMutex.NewNSLock(bucket, object)
 	if err := objectLock.GetLock(globalOperationTimeout); err != nil {
@@ -825,7 +825,7 @@ func (fs *fsObjects) DeleteObject(bucket, object string) error {
 // Returns function "listDir" of the type listDirFunc.
 // isLeaf - is used by listDir function to check if an entry
 // is a leaf or non-leaf entry.
-func (fs *fsObjects) listDirFactory(isLeaf isLeafFunc) listDirFunc {
+func (fs *FSObjects) listDirFactory(isLeaf isLeafFunc) listDirFunc {
 	// listDir - lists all the entries at a given prefix and given entry in the prefix.
 	listDir := func(bucket, prefixDir, prefixEntry string) (entries []string, delayIsLeaf bool, err error) {
 		entries, err = readDir(pathJoin(fs.fsPath, bucket, prefixDir))
@@ -842,7 +842,7 @@ func (fs *fsObjects) listDirFactory(isLeaf isLeafFunc) listDirFunc {
 
 // getObjectETag is a helper function, which returns only the md5sum
 // of the file on the disk.
-func (fs *fsObjects) getObjectETag(bucket, entry string) (string, error) {
+func (fs *FSObjects) getObjectETag(bucket, entry string) (string, error) {
 	fsMetaPath := pathJoin(fs.fsPath, minioMetaBucket, bucketMetaPrefix, bucket, entry, fsMetaJSONFile)
 
 	// Read `fs.json` to perhaps contend with
@@ -891,7 +891,7 @@ func (fs *fsObjects) getObjectETag(bucket, entry string) (string, error) {
 
 // ListObjects - list all objects at prefix upto maxKeys., optionally delimited by '/'. Maintains the list pool
 // state for future re-entrant list requests.
-func (fs *fsObjects) ListObjects(bucket, prefix, marker, delimiter string, maxKeys int) (loi ListObjectsInfo, e error) {
+func (fs *FSObjects) ListObjects(bucket, prefix, marker, delimiter string, maxKeys int) (loi ListObjectsInfo, e error) {
 	if err := checkListObjsArgs(bucket, prefix, marker, delimiter, fs); err != nil {
 		return loi, err
 	}
@@ -1051,53 +1051,53 @@ func (fs *fsObjects) ListObjects(bucket, prefix, marker, delimiter string, maxKe
 }
 
 // HealFormat - no-op for fs, Valid only for XL.
-func (fs *fsObjects) HealFormat(dryRun bool) (madmin.HealResultItem, error) {
+func (fs *FSObjects) HealFormat(dryRun bool) (madmin.HealResultItem, error) {
 	return madmin.HealResultItem{}, errors.Trace(NotImplemented{})
 }
 
 // HealObject - no-op for fs. Valid only for XL.
-func (fs *fsObjects) HealObject(bucket, object string, dryRun bool) (
+func (fs *FSObjects) HealObject(bucket, object string, dryRun bool) (
 	res madmin.HealResultItem, err error) {
 	return res, errors.Trace(NotImplemented{})
 }
 
 // HealBucket - no-op for fs, Valid only for XL.
-func (fs *fsObjects) HealBucket(bucket string, dryRun bool) ([]madmin.HealResultItem,
+func (fs *FSObjects) HealBucket(bucket string, dryRun bool) ([]madmin.HealResultItem,
 	error) {
 	return nil, errors.Trace(NotImplemented{})
 }
 
 // ListObjectsHeal - list all objects to be healed. Valid only for XL
-func (fs *fsObjects) ListObjectsHeal(bucket, prefix, marker, delimiter string, maxKeys int) (loi ListObjectsInfo, e error) {
+func (fs *FSObjects) ListObjectsHeal(bucket, prefix, marker, delimiter string, maxKeys int) (loi ListObjectsInfo, e error) {
 	return loi, errors.Trace(NotImplemented{})
 }
 
 // ListBucketsHeal - list all buckets to be healed. Valid only for XL
-func (fs *fsObjects) ListBucketsHeal() ([]BucketInfo, error) {
+func (fs *FSObjects) ListBucketsHeal() ([]BucketInfo, error) {
 	return []BucketInfo{}, errors.Trace(NotImplemented{})
 }
 
 // SetBucketPolicy sets policy on bucket
-func (fs *fsObjects) SetBucketPolicy(bucket string, policy policy.BucketAccessPolicy) error {
+func (fs *FSObjects) SetBucketPolicy(bucket string, policy policy.BucketAccessPolicy) error {
 	return persistAndNotifyBucketPolicyChange(bucket, false, policy, fs)
 }
 
 // GetBucketPolicy will get policy on bucket
-func (fs *fsObjects) GetBucketPolicy(bucket string) (policy.BucketAccessPolicy, error) {
+func (fs *FSObjects) GetBucketPolicy(bucket string) (policy.BucketAccessPolicy, error) {
 	policy := fs.bucketPolicies.GetBucketPolicy(bucket)
 	if reflect.DeepEqual(policy, emptyBucketPolicy) {
-		return readBucketPolicy(bucket, fs)
+		return ReadBucketPolicy(bucket, fs)
 	}
 	return policy, nil
 }
 
 // DeleteBucketPolicy deletes all policies on bucket
-func (fs *fsObjects) DeleteBucketPolicy(bucket string) error {
+func (fs *FSObjects) DeleteBucketPolicy(bucket string) error {
 	return persistAndNotifyBucketPolicyChange(bucket, true, emptyBucketPolicy, fs)
 }
 
 // ListObjectsV2 lists all blobs in bucket filtered by prefix
-func (fs *fsObjects) ListObjectsV2(bucket, prefix, continuationToken, delimiter string, maxKeys int, fetchOwner bool, startAfter string) (result ListObjectsV2Info, err error) {
+func (fs *FSObjects) ListObjectsV2(bucket, prefix, continuationToken, delimiter string, maxKeys int, fetchOwner bool, startAfter string) (result ListObjectsV2Info, err error) {
 	loi, err := fs.ListObjects(bucket, prefix, continuationToken, delimiter, maxKeys)
 	if err != nil {
 		return result, err
@@ -1114,8 +1114,8 @@ func (fs *fsObjects) ListObjectsV2(bucket, prefix, continuationToken, delimiter
 }
 
 // RefreshBucketPolicy refreshes cache policy with what's on disk.
-func (fs *fsObjects) RefreshBucketPolicy(bucket string) error {
-	policy, err := readBucketPolicy(bucket, fs)
+func (fs *FSObjects) RefreshBucketPolicy(bucket string) error {
+	policy, err := ReadBucketPolicy(bucket, fs)
 
 	if err != nil {
 		if reflect.DeepEqual(policy, emptyBucketPolicy) {
@@ -1127,11 +1127,11 @@ func (fs *fsObjects) RefreshBucketPolicy(bucket string) error {
 }
 
 // IsNotificationSupported returns whether bucket notification is applicable for this layer.
-func (fs *fsObjects) IsNotificationSupported() bool {
+func (fs *FSObjects) IsNotificationSupported() bool {
 	return true
 }
 
 // IsEncryptionSupported returns whether server side encryption is applicable for this layer.
-func (fs *fsObjects) IsEncryptionSupported() bool {
+func (fs *FSObjects) IsEncryptionSupported() bool {
 	return true
 }
diff --git a/cmd/fs-v1_test.go b/cmd/fs-v1_test.go
index 2f6c7d47d..e8f849341 100644
--- a/cmd/fs-v1_test.go
+++ b/cmd/fs-v1_test.go
@@ -56,7 +56,7 @@ func TestFSParentDirIsObject(t *testing.T) {
 		t.Fatalf("Unexpected object name returned got %s, expected %s", objInfo.Name, objectName)
 	}
 
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	testCases := []struct {
 		parentIsObject bool
 		objectName     string
@@ -101,16 +101,16 @@ func TestFSParentDirIsObject(t *testing.T) {
 // and constructs a valid `FS` object layer.
 func TestNewFS(t *testing.T) {
 	// Do not attempt to create this path, the test validates
-	// so that newFSObjectLayer initializes non existing paths
+	// so that NewFSObjectLayer initializes non existing paths
 	// and successfully returns initialized object layer.
 	disk := filepath.Join(globalTestTmpDir, "minio-"+nextSuffix())
 	defer os.RemoveAll(disk)
 
-	_, err := newFSObjectLayer("")
+	_, err := NewFSObjectLayer("")
 	if err != errInvalidArgument {
 		t.Errorf("Expecting error invalid argument, got %s", err)
 	}
-	_, err = newFSObjectLayer(disk)
+	_, err = NewFSObjectLayer(disk)
 	if err != nil {
 		errMsg := "Unable to recognize backend format, Disk is not in FS format."
 		if err.Error() == errMsg {
@@ -131,10 +131,10 @@ func TestFSShutdown(t *testing.T) {
 	bucketName := "testbucket"
 	objectName := "object"
 	// Create and return an fsObject with its path in the disk
-	prepareTest := func() (*fsObjects, string) {
+	prepareTest := func() (*FSObjects, string) {
 		disk := filepath.Join(globalTestTmpDir, "minio-"+nextSuffix())
 		obj := initFSObjects(disk, t)
-		fs := obj.(*fsObjects)
+		fs := obj.(*FSObjects)
 		objectContent := "12345"
 		obj.MakeBucketWithLocation(bucketName, "")
 		obj.PutObject(bucketName, objectName, mustGetHashReader(t, bytes.NewReader([]byte(objectContent)), int64(len(objectContent)), "", ""), nil)
@@ -164,7 +164,7 @@ func TestFSGetBucketInfo(t *testing.T) {
 	defer os.RemoveAll(disk)
 
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 
 	obj.MakeBucketWithLocation(bucketName, "")
@@ -266,7 +266,7 @@ func TestFSDeleteObject(t *testing.T) {
 	defer os.RemoveAll(disk)
 
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 	objectName := "object"
 
@@ -311,7 +311,7 @@ func TestFSDeleteBucket(t *testing.T) {
 	defer os.RemoveAll(disk)
 
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 	bucketName := "bucket"
 
 	err := obj.MakeBucketWithLocation(bucketName, "")
@@ -350,7 +350,7 @@ func TestFSListBuckets(t *testing.T) {
 	defer os.RemoveAll(disk)
 
 	obj := initFSObjects(disk, t)
-	fs := obj.(*fsObjects)
+	fs := obj.(*FSObjects)
 
 	bucketName := "bucket"
 	if err := obj.MakeBucketWithLocation(bucketName, ""); err != nil {
diff --git a/cmd/gateway/gateway.go b/cmd/gateway/gateway.go
index 4454a5e52..656f56c4b 100644
--- a/cmd/gateway/gateway.go
+++ b/cmd/gateway/gateway.go
@@ -22,6 +22,7 @@ import (
 	_ "github.com/minio/minio/cmd/gateway/b2"
 	_ "github.com/minio/minio/cmd/gateway/gcs"
 	_ "github.com/minio/minio/cmd/gateway/manta"
+	_ "github.com/minio/minio/cmd/gateway/nas"
 	_ "github.com/minio/minio/cmd/gateway/oss"
 	_ "github.com/minio/minio/cmd/gateway/s3"
 	_ "github.com/minio/minio/cmd/gateway/sia"
diff --git a/cmd/gateway/nas/gateway-nas.go b/cmd/gateway/nas/gateway-nas.go
new file mode 100644
index 000000000..7c16796c1
--- /dev/null
+++ b/cmd/gateway/nas/gateway-nas.go
@@ -0,0 +1,119 @@
+/*
+ * Minio Cloud Storage, (C) 2018 Minio, Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package nas
+
+import (
+	"github.com/minio/cli"
+	"github.com/minio/minio-go/pkg/policy"
+	minio "github.com/minio/minio/cmd"
+	"github.com/minio/minio/pkg/auth"
+)
+
+const (
+	nasBackend = "nas"
+)
+
+func init() {
+	const nasGatewayTemplate = `NAME:
+  {{.HelpName}} - {{.Usage}}
+
+USAGE:
+  {{.HelpName}} {{if .VisibleFlags}}[FLAGS]{{end}} PATH
+{{if .VisibleFlags}}
+FLAGS:
+  {{range .VisibleFlags}}{{.}}
+  {{end}}{{end}}
+PATH:
+  path to NAS mount point.
+
+ENVIRONMENT VARIABLES:
+  ACCESS:
+     MINIO_ACCESS_KEY: Username or access key of minimum 3 characters in length.
+     MINIO_SECRET_KEY: Password or secret key of minimum 8 characters in length.
+
+  BROWSER:
+     MINIO_BROWSER: To disable web browser access, set this value to "off".
+
+  UPDATE:
+     MINIO_UPDATE: To turn off in-place upgrades, set this value to "off".
+
+EXAMPLES:
+  1. Start minio gateway server for NAS backend.
+      $ export MINIO_ACCESS_KEY=accesskey
+      $ export MINIO_SECRET_KEY=secretkey
+      $ {{.HelpName}} /shared/nasvol
+`
+
+	minio.RegisterGatewayCommand(cli.Command{
+		Name:               nasBackend,
+		Usage:              "Network-attached storage (NAS).",
+		Action:             nasGatewayMain,
+		CustomHelpTemplate: nasGatewayTemplate,
+		HideHelpCommand:    true,
+	})
+}
+
+// Handler for 'minio gateway nas' command line.
+func nasGatewayMain(ctx *cli.Context) {
+	// Validate gateway arguments.
+	host := ctx.Args().First()
+	if host == "" {
+		cli.ShowCommandHelpAndExit(ctx, "nas", 1)
+	}
+	// Validate gateway arguments.
+	minio.StartGateway(ctx, &NAS{host})
+}
+
+// NAS implements Gateway.
+type NAS struct {
+	host string
+}
+
+// Name implements Gateway interface.
+func (g *NAS) Name() string {
+	return nasBackend
+}
+
+// NewGatewayLayer returns nas gatewaylayer.
+func (g *NAS) NewGatewayLayer(creds auth.Credentials) (minio.ObjectLayer, error) {
+	var err error
+	newObject, err := minio.NewFSObjectLayer(g.host)
+	if err != nil {
+		return nil, err
+	}
+	return &nasObjects{newObject.(*minio.FSObjects)}, nil
+}
+
+// Production - nas gateway is production ready.
+func (g *NAS) Production() bool {
+	return true
+}
+
+// nasObjects implements gateway for Minio and S3 compatible object storage servers.
+type nasObjects struct {
+	*minio.FSObjects
+}
+
+// IsNotificationSupported returns whether notifications are applicable for this layer.
+func (l *nasObjects) IsNotificationSupported() bool {
+	return false
+}
+
+// GetBucketPolicy will get policy on bucket
+func (l *nasObjects) GetBucketPolicy(bucket string) (policy.BucketAccessPolicy, error) {
+	return minio.ReadBucketPolicy(bucket, l)
+}
diff --git a/cmd/object-api-listobjects_test.go b/cmd/object-api-listobjects_test.go
index 2344c26e2..409f13287 100644
--- a/cmd/object-api-listobjects_test.go
+++ b/cmd/object-api-listobjects_test.go
@@ -576,7 +576,7 @@ func testListObjects(obj ObjectLayer, instanceType string, t TestErrHandler) {
 // Initialize FS backend for the benchmark.
 func initFSObjectsB(disk string, t *testing.B) (obj ObjectLayer) {
 	var err error
-	obj, err = newFSObjectLayer(disk)
+	obj, err = NewFSObjectLayer(disk)
 	if err != nil {
 		t.Fatal("Unexpected err: ", err)
 	}
diff --git a/cmd/server-main.go b/cmd/server-main.go
index de3e33215..091f153c1 100644
--- a/cmd/server-main.go
+++ b/cmd/server-main.go
@@ -270,7 +270,7 @@ func newObjectLayer(endpoints EndpointList) (newObject ObjectLayer, err error) {
 	isFS := len(endpoints) == 1
 	if isFS {
 		// Initialize new FS object layer.
-		return newFSObjectLayer(endpoints[0].Path)
+		return NewFSObjectLayer(endpoints[0].Path)
 	}
 
 	format, err := waitForFormatXL(endpoints[0].IsLocal, endpoints, globalXLSetCount, globalXLSetDriveCount)
diff --git a/cmd/server-main_test.go b/cmd/server-main_test.go
index 1334ba152..663090407 100644
--- a/cmd/server-main_test.go
+++ b/cmd/server-main_test.go
@@ -36,7 +36,7 @@ func TestNewObjectLayer(t *testing.T) {
 	if err != nil {
 		t.Fatal("Unexpected object layer initialization error", err)
 	}
-	_, ok := obj.(*fsObjects)
+	_, ok := obj.(*FSObjects)
 	if !ok {
 		t.Fatal("Unexpected object layer detected", reflect.TypeOf(obj))
 	}
diff --git a/cmd/test-utils_test.go b/cmd/test-utils_test.go
index d624dfba6..b9c27d4db 100644
--- a/cmd/test-utils_test.go
+++ b/cmd/test-utils_test.go
@@ -163,7 +163,7 @@ func prepareFS() (ObjectLayer, string, error) {
 	if err != nil {
 		return nil, "", err
 	}
-	obj, err := newFSObjectLayer(fsDirs[0])
+	obj, err := NewFSObjectLayer(fsDirs[0])
 	if err != nil {
 		return nil, "", err
 	}
@@ -221,7 +221,7 @@ func prepareXL16() (ObjectLayer, []string, error) {
 func initFSObjects(disk string, t *testing.T) (obj ObjectLayer) {
 	newTestConfig(globalMinioDefaultRegion)
 	var err error
-	obj, err = newFSObjectLayer(disk)
+	obj, err = NewFSObjectLayer(disk)
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -1685,7 +1685,7 @@ func newTestObjectLayer(endpoints EndpointList) (newObject ObjectLayer, err erro
 	isFS := len(endpoints) == 1
 	if isFS {
 		// Initialize new FS object layer.
-		return newFSObjectLayer(endpoints[0].Path)
+		return NewFSObjectLayer(endpoints[0].Path)
 	}
 
 	_, err = waitForFormatXL(endpoints[0].IsLocal, endpoints, 1, 16)
diff --git a/cmd/xl-sets.go b/cmd/xl-sets.go
index fb784f448..8144df8e8 100644
--- a/cmd/xl-sets.go
+++ b/cmd/xl-sets.go
@@ -433,7 +433,7 @@ func (s *xlSets) GetBucketPolicy(bucket string) (policy.BucketAccessPolicy, erro
 	// fetch bucket policy from cache.
 	bpolicy := s.bucketPolicies.GetBucketPolicy(bucket)
 	if reflect.DeepEqual(bpolicy, emptyBucketPolicy) {
-		return readBucketPolicy(bucket, s)
+		return ReadBucketPolicy(bucket, s)
 	}
 	return bpolicy, nil
 }
@@ -445,7 +445,7 @@ func (s *xlSets) DeleteBucketPolicy(bucket string) error {
 
 // RefreshBucketPolicy refreshes policy cache from disk
 func (s *xlSets) RefreshBucketPolicy(bucket string) error {
-	policy, err := readBucketPolicy(bucket, s)
+	policy, err := ReadBucketPolicy(bucket, s)
 	if err != nil {
 		if reflect.DeepEqual(policy, emptyBucketPolicy) {
 			return s.bucketPolicies.DeleteBucketPolicy(bucket)
diff --git a/cmd/xl-v1-bucket.go b/cmd/xl-v1-bucket.go
index 54197d8e1..f36e295ae 100644
--- a/cmd/xl-v1-bucket.go
+++ b/cmd/xl-v1-bucket.go
@@ -291,7 +291,7 @@ func (xl xlObjects) GetBucketPolicy(bucket string) (policy.BucketAccessPolicy, e
 	// fetch bucket policy from cache.
 	bpolicy := xl.bucketPolicies.GetBucketPolicy(bucket)
 	if reflect.DeepEqual(bpolicy, emptyBucketPolicy) {
-		return readBucketPolicy(bucket, xl)
+		return ReadBucketPolicy(bucket, xl)
 	}
 	return bpolicy, nil
 }
@@ -303,7 +303,7 @@ func (xl xlObjects) DeleteBucketPolicy(bucket string) error {
 
 // RefreshBucketPolicy refreshes policy cache from disk
 func (xl xlObjects) RefreshBucketPolicy(bucket string) error {
-	policy, err := readBucketPolicy(bucket, xl)
+	policy, err := ReadBucketPolicy(bucket, xl)
 
 	if err != nil {
 		if reflect.DeepEqual(policy, emptyBucketPolicy) {
diff --git a/docs/gateway/azure.md b/docs/gateway/azure.md
index 3ef710ca0..9e915a477 100644
--- a/docs/gateway/azure.md
+++ b/docs/gateway/azure.md
@@ -17,7 +17,7 @@ export MINIO_SECRET_KEY=azureaccountkey
 minio gateway azure
 ```
 ## Test using Minio Browser
-Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 ensure your server has started successfully.
+Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 to ensure that your server has started successfully.
 
 ![Screenshot](https://github.com/minio/minio/blob/master/docs/screenshots/minio-browser-gateway.png?raw=true)
 ## Test using Minio Client `mc`
diff --git a/docs/gateway/b2.md b/docs/gateway/b2.md
index f413be4d2..b0ed41cb1 100644
--- a/docs/gateway/b2.md
+++ b/docs/gateway/b2.md
@@ -13,7 +13,7 @@ docker run -p 9000:9000 --name b2-s3 \
 ```
 
 ## Test using Minio Browser
-Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 ensure your server has started successfully.
+Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 to ensure that your server has started successfully.
 
 ![Screenshot](https://raw.githubusercontent.com/minio/minio/master/docs/screenshots/minio-browser-gateway.png)
 
diff --git a/docs/gateway/gcs.md b/docs/gateway/gcs.md
index cbb825b21..20bc331df 100644
--- a/docs/gateway/gcs.md
+++ b/docs/gateway/gcs.md
@@ -32,7 +32,7 @@ minio gateway gcs yourprojectid
 ```
 
 ## Test using Minio Browser
-Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 ensure your server has started successfully.
+Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 to ensure that your server has started successfully.
 
 ![Screenshot](https://github.com/minio/minio/blob/master/docs/screenshots/minio-browser-gateway.png?raw=true)
 
diff --git a/docs/gateway/manta.md b/docs/gateway/manta.md
index 70d745613..e3073f0a0 100644
--- a/docs/gateway/manta.md
+++ b/docs/gateway/manta.md
@@ -21,7 +21,7 @@ export MANTA_SUBUSER=devuser
 minio gateway manta
 ```
 ## Test using Minio Browser
-Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 ensure your server has started successfully.
+Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 to ensure that your server has started successfully.
 
 ![Screenshot](https://github.com/minio/minio/blob/master/docs/screenshots/minio-browser-gateway.png?raw=true)
 ## Test using Minio Client `mc`
diff --git a/docs/gateway/nas.md b/docs/gateway/nas.md
new file mode 100644
index 000000000..724def924
--- /dev/null
+++ b/docs/gateway/nas.md
@@ -0,0 +1,43 @@
+# Minio NAS Gateway [![Slack](https://slack.minio.io/slack?type=svg)](https://slack.minio.io)
+Minio Gateway adds Amazon S3 compatibility to NAS storage. You may run multiple minio instances on the same shared NAS volume as a distributed object gateway.
+
+## Run Minio Gateway for NAS Storage
+### Using Docker
+```
+docker run -p 9000:9000 --name nas-s3 \
+ -e "MINIO_ACCESS_KEY=minio" \
+ -e "MINIO_SECRET_KEY=minio123" \
+ minio/minio:edge gateway nas /shared/nasvol
+```
+
+### Using Binary
+```
+export MINIO_ACCESS_KEY=minioaccesskey
+export MINIO_SECRET_KEY=miniosecretkey
+minio gateway nas /shared/nasvol
+```
+## Test using Minio Browser
+Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 to ensure that your server has started successfully.
+
+![Screenshot](https://raw.githubusercontent.com/minio/minio/master/docs/screenshots/minio-browser-gateway.png)
+
+## Test using Minio Client `mc`
+`mc` provides a modern alternative to UNIX commands such as ls, cat, cp, mirror, diff etc. It supports filesystems and Amazon S3 compatible cloud storage services.
+
+### Configure `mc`
+```
+mc config host add mynas http://gateway-ip:9000 access_key secret_key
+```
+
+### List buckets on nas
+```
+mc ls mynas
+[2017-02-22 01:50:43 PST]     0B ferenginar/
+[2017-02-26 21:43:51 PST]     0B my-bucket/
+[2017-02-26 22:10:11 PST]     0B test-bucket1/
+```
+
+## Explore Further
+- [`mc` command-line interface](https://docs.minio.io/docs/minio-client-quickstart-guide)
+- [`aws` command-line interface](https://docs.minio.io/docs/aws-cli-with-minio)
+- [`minio-go` Go SDK](https://docs.minio.io/docs/golang-client-quickstart-guide)
diff --git a/docs/gateway/oss.md b/docs/gateway/oss.md
index 77dbcc998..015e143c1 100644
--- a/docs/gateway/oss.md
+++ b/docs/gateway/oss.md
@@ -19,7 +19,7 @@ minio gateway azure
 ```
 
 ## Test using Minio Browser
-Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 ensure your server has started successfully.
+Minio Gateway comes with an embedded web based object browser. Point your web browser to http://127.0.0.1:9000 to ensure that your server has started successfully.
 
 ![Screenshot](https://raw.githubusercontent.com/minio/minio/master/docs/screenshots/minio-browser-gateway.png)