This document uses the following terms:
8.3 name: A file name string restricted in length to 12 characters that includes a base name of up to eight characters, one character for a period, and up to three characters for a file name extension. For more information on 8.3 file names, see [MS-CIFS] section 18.104.22.168.1.
alternate name: An 8.3 name that can optionally be generated when a file is created. A file will not have an alternate name if the user wants to optimize performance, or if the name of the file already uses the 8.3 format.
chunk: The amount of data that the operating system's implementation of the Lempel-Ziv compression algorithm tries to compress at one time. The compression unit size used by the file system is always a multiple of the underlying compression algorithm's chunk size. For more information on the Lempel-Ziv compression algorithm, see [UASDC].
cluster: The smallest allocation unit on a volume.
compression unit: The amount of data that NTFS tries to compress at one time. Compression of large files is accomplished as a series of compressions of data blocks, each at the most compression unit bytes in size.
compression unit shift: The number of bits by which to left-shift a 1 bit to arrive at the compression unit size.
disk quota: Maximum amount of data a user can store on a disk volume.
Distributed Link Tracking (DLT): A protocol that enables client applications to track sources that have been sent to remote locations using remote procedure call (RPC) interfaces, and to maintain links to files. It exposes methods that belong to two interfaces, one of which exists on the server (trksvr) and the other on a workstation (trkwks).
FAT file system: A file system used by MS-DOS and other Windows operating systems to organize and manage files. The file allocation table (FAT) is a data structure that the operating system creates when a volume is formatted by using FAT or FAT32 file systems. The operating system stores information about each file in the FAT so that it can retrieve the file later.
Fid: A 16-bit value that the Server Message Block (SMB) server uses to represent an opened file, named pipe, printer, or device. A Fid is returned by an SMB server in response to a client request to open or create a file, named pipe, printer, or device. The SMB server guarantees that the Fid value returned is unique for a given SMB connection until the SMB connection is closed, at which time the Fid value can be reused. The Fid is used by the SMB client in subsequent SMB commands to identify the opened file, named pipe, printer, or device.
file allocation table (FAT): A data structure that the operating system creates when a volume is formatted by using FAT or FAT32 file systems. The operating system stores information about each file in the FAT so that it can retrieve the file later.
file stream: See main stream and named stream.
file system control (FSCTL): A command issued to a file system to alter or query the behavior of the file system and/or set or query metadata that is associated with a particular file or with the file system itself.
filter: Type of driver that is layered between the kernel and a base file system (such as FAT or NTFS) that receives I/O request packets on their way to and from the base file system. The term filter can refer to legacy filters or minifilters.
filter manager: A file system filter driver that simplifies the development of other file system filter drivers. Although it is possible to write a filter driver that manages other filters, for the purposes of this document, the phrase filter manager refers only to the file system filter manager, which is an operating system component. A filter driver developed to the filter manager model is called a minifilter.
globally unique identifier (GUID): A term used interchangeably with universally unique identifier (UUID) in Microsoft protocol technical documents (TDs). Interchanging the usage of these terms does not imply or require a specific algorithm or mechanism to generate the value. Specifically, the use of this term does not imply or require that the algorithms described in [RFC4122] or [C706] must be used for generating the GUID. See also universally unique identifier (UUID).
GUIDString: A GUID in the form of an ASCII or Unicode string, consisting of one group of 8 hexadecimal digits, followed by three groups of 4 hexadecimal digits each, followed by one group of 12 hexadecimal digits. It is the standard representation of a GUID, as described in [RFC4122] section 3. For example, "6B29FC40-CA47-1067-B31D-00DD010662DA". Unlike a curly braced GUID string, a GUIDString is not enclosed in braces.
master file table (MFT): On an NTFS volume, the MFT is a relational database that consists of rows of file records and columns of file attributes. It contains at least one entry for every file on an NTFS volume, including the MFT itself. The MFT stores the information required to retrieve files from the NTFS partition.
named stream: A place within a file in addition to the main stream where data is stored, or the data stored therein. File systems support a mode in which it is possible to open either the main stream of a file and/or to open a named stream. Named streams have different data than the main stream (and than each other) and can be read and written independently. Not all file systems support named streams. See also main stream.
NT file system (NTFS): A proprietary Microsoft file system. For more information, see [MSFT-NTFS].
Offload Read: A variant to a normal read operation where a target device generates and returns a Token instead of a buffer containing the data to be read. The Token is maintained by the target device until it invalidates the Token for any vendor-specific reason. The data logically represented by the Token cannot change, and the target device is required to maintain this representation. An example of a target device is a SAN Storage Array with support for the associated low-level storage commands. For more information on Offload Read, see [INCITS-T10/11-059].
Offload Write: A variant to a normal write operation where the host provides a Token instead of a buffer containing the data to be written. Upon receipt of the Offload Write, the target device parses the Token and determines whether the data movement (the Write) can be completed to the requested location. An example of a target device is a SAN Storage Array with support for the associated low-level storage commands. For more information on Offload Write, see [INCITS-T10/11-059].
reparse point: An attribute that can be added to a file to store a collection of user-defined data that is opaque to NTFS or ReFS. If a file that has a reparse point is opened, the open will normally fail with STATUS_REPARSE, so that the relevant file system filter driver can detect the open of a file associated with (owned by) this reparse point. At that point, each installed filter driver can check to see if it is the owner of the reparse point, and, if so, perform any special processing required for a file with that reparse point. The format of this data is understood by the application that stores the data and the file system filter that interprets the data and processes the file. For example, an encryption filter that is marked as the owner of a file's reparse point could look up the encryption key for that file. A file can have (at most) 1 reparse point associated with it. For more information, see [MS-FSCC].
reparse point tag: A unique identifier for a file system filter driver stored within a file's optional reparse point data that indicates the file system filter driver that performs additional filter-defined processing on a file during I/O operations. An implementer can request more than one reparse point for use with a file system, a file system filter driver, or a minifilter driver. To request a reparse point tag, use the reparse point tag request form. For more information, see [WHDC-RPTR].
replica set: In File Replication Service (FRS), the replication of files and directories according to a predefined topology and schedule on a specific folder. The topology and schedule are collectively called a replica set. A replica set contains a set of replicas, one for each machine that participates in replication.
security identifier (SID): An identifier for security principals in Windows that is used to identify an account or a group. Conceptually, the SID is composed of an account authority portion (typically a domain) and a smaller integer representing an identity relative to the account authority, termed the relative identifier (RID). The SID format is specified in [MS-DTYP] section 2.4.2; a string representation of SIDs is specified in [MS-DTYP] section 2.4.2 and [MS-AZOD] section 22.214.171.124.
short name: This has the same definition as alternate name.
single-instance storage (SIS): An NTFS feature that implements links with the semantics of copies for files stored on an NTFS volume. SIS uses copy-on-close to implement the copy semantics of its links.
sparse file: A file containing large sections of data composed only of zeros, which is marked as such in the NTFS. The file system saves disk space by only allocating as many ranges on disk as are required to completely reconstruct the non-zero data. When an attempt is made to read in the nonallocated portions of the file (also known as holes), the file system automatically returns zeros to the caller.
stream: A sequence of bytes written to a file on the NTFS file system. Every file stored on a volume that uses the NTFS file system contains at least one stream, which is normally used to store the primary contents of the file. Additional streams within the file can be used to store file attributes, application parameters, or other information specific to that file. Every file has a default data stream, which is unnamed by default. That data stream, and any other data stream associated with a file, can optionally be named.
sub-read and sub-write: An I/O operation sent by the file system to the storage stack that is part of a larger file I/O operation. Sometimes large file reads and writes are broken down by the file system into smaller reads and writes, which are then sent to the storage stack.
symbolic link: A symbolic link is a reparse point that points to another file system object. The object being pointed to is called the target. Symbolic links are transparent to users; the links appear as normal files or directories, and can be acted upon by the user or application in exactly the same manner. Symbolic links can be created using the FSCTL_SET_REPARSE_POINT request as specified in [MS-FSCC] section 2.3.61. They can be deleted using the FSCTL_DELETE_REPARSE_POINT request as specified in [MS-FSCC] section 2.3.5. Implementing symbolic links is optional for a file system.
tag: Another name for a reparse point. For instance, the file system filter manager FltTagFile routine sets a reparse point on a file. Tag is also used to refer to the field in a reparse point that identifies what software component put the reparse point there.
token: A 512-byte length opaque string that is generated and maintained by a supported target device. A Token functions logically as an immutable point-in-time representation for a set of data specified by a host and can be conceptualized as a compressed representation of the data that only a certain class of storage subsystems can interpret. A Token can also be constructed from a set of well-known Tokens to enable the client to describe a homogeneous attribute for a set of data (for example, all zeros) or to enable a server to apply a homogeneous attribute to a set of data (for example, a set of all zeros). For more information on Tokens, see [INCITS-T10/11-059].
Uniform Resource Locator (URL): A string of characters in a standardized format that identifies a document or resource on the World Wide Web. The format is as specified in [RFC1738].
volume: A group of one or more partitions that forms a logical region of storage and the basis for a file system. A volume is an area on a storage device that is managed by the file system as a discrete logical storage unit. A partition contains at least one volume, and a volume can exist on one or more partitions.
MAY, SHOULD, MUST, SHOULD NOT, MUST NOT: These terms (in all caps) are used as defined in [RFC2119]. All statements of optional behavior use either MAY, SHOULD, or SHOULD NOT.