INTERNET-DRAFT                                     Christopher R. Hertel
draft-crhertel-smb-url-12.txt                                 Samba Team
Expires July 8, 2007                                     January 8, 2007

                      SMB File Sharing URI Scheme

Intellectual Property Rights Disclosure Acknowledgment

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

Status of this Memo

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups. Note that other
   groups may also distribute working documents as Internet-Drafts.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time. It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at

   The list of Internet-Draft Shadow Directories can be accessed at

   Discussions regarding this document and the SMB URI scheme should
   take place on the mailing list.

   This Internet-Draft will expire on July 8th, 2007.


   The Server Message Block (SMB) protocol is one of the most widely
   used network file system protocols in existence.  This document
   describes a scheme for an SMB Uniform Resource Identifier (SMB URI)
   which can be used to identify SMB workgroups, servers, server shares,
   directories, files, inter-process communications ports (named pipes),
   and devices--the objects in the SMB network file system space.

Hertel                     Expires July 8, 2007                 [Page 1]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

Table of Contents

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
   1.1. Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. SMB URI Scheme Overview . . . . . . . . . . . . . . . . . . . . . . 4
3. NetBIOS, NBT, and NBT Transport Semantics . . . . . . . . . . . . . 4
   3.1. The NetBIOS Name Service . . . . . . . . . . . . . . . . . . . 5
   3.2. The NetBIOS Datagram Service . . . . . . . . . . . . . . . . . 5
   3.3. The NetBIOS Session Service. . . . . . . . . . . . . . . . . . 5
   3.4  Accommodating NBT in the SMB URI . . . . . . . . . . . . . . . 5
4. NetBIOS-Based Workgroups. . . . . . . . . . . . . . . . . . . . . . 6
5. SMB URI Definition. . . . . . . . . . . . . . . . . . . . . . . . . 9
6. SMB URI Syntax Elements . . . . . . . . . . . . . . . . . . . . . .10
   6.1. scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
   6.2. smb-service. . . . . . . . . . . . . . . . . . . . . . . . . .10
   6.3. auth-domain. . . . . . . . . . . . . . . . . . . . . . . . . .11
   6.4. smb-srv-name . . . . . . . . . . . . . . . . . . . . . . . . .11
   6.5. port . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
   6.6. scope-id . . . . . . . . . . . . . . . . . . . . . . . . . . .12
   6.7. nbt-context. . . . . . . . . . . . . . . . . . . . . . . . . .13
7. The Relationship Between the SMB URI and the UNC Format . . . . . .15
8. Authentication. . . . . . . . . . . . . . . . . . . . . . . . . . .16
9. Security Considerations . . . . . . . . . . . . . . . . . . . . . .16
10. Character Encoding Issues. . . . . . . . . . . . . . . . . . . . .17
11. Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . .17
12. References . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
13. Author's Address . . . . . . . . . . . . . . . . . . . . . . . . .18
14. Copyright Notice . . . . . . . . . . . . . . . . . . . . . . . . .19
15. Disclaimer of Validity . . . . . . . . . . . . . . . . . . . . . .19
Appendix A. Working with NetBIOS Names (Implementation Notes). . . . .19
   A.1. NetBIOS Names. . . . . . . . . . . . . . . . . . . . . . . . .19
   A.2. SMB Sessions via NBT . . . . . . . . . . . . . . . . . . . . .19
   A.3. Resolving DNS names and IP addresses to SMB server names . . .20
   A.4. Determining the Namespace of the smb-srv-name. . . . . . . . .22
   A.5. Workgroup vs. SMB Server Names . . . . . . . . . . . . . . . .23

Hertel                     Expires July 8, 2007                 [Page 2]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

1. Introduction

   The Server Message Block protocol (SMB) was created in the 1980's by
   Dr. Barry Feigenbaum at IBM Corporation.  It was later extended by
   various contributors at 3Com, IBM, Intel, and Microsoft.  During the
   mid 1990's SMB was renamed CIFS (Common Internet File System).  Both
   names are in general use today.  In common parlance, however, the
   term "SMB" is typically used when referring to the core filesharing
   protocol itself, while "CIFS" includes the set of sub-protocols and
   extensions used to create the complete filesharing suite.

   SMB was originally carried via a proprietary network transport, the
   interface to which was called NetBIOS (Network Basic Input Output
   System).  Two Internet RFCs ([RFC1001], [RFC1002]) were published
   which describe a mechanism for implementing the NetBIOS API on top
   of TCP and UDP, thus allowing SMB to be transported over IP inter-
   networks.  Those RFCs are now known collectively as Internet
   Standard #19, and the transport protocol that they describe is
   commonly called NBT (or, sometimes, NetBT) for "NetBIOS over

   In addition to transport via NBT, newer implementations of SMB
   typically support SMB over TCP/IP without the intervening NetBIOS
   emulation layer.  This is known as SMB over "native TCP" or "naked

   Several attempts have been made to document and even standardize the
   CIFS suite ([XOPENSMB], [ONET], [SNIACIFS], [IMPCIFS]), yet the
   further development of CIFS remains under the control of Microsoft.
   Despite its proprietary nature, the workings of SMB are sufficiently
   well known that SMB file sharing has been successfully implemented
   by several third-party commercial vendors and in Open Source.  SMB
   server and client software is available for a wide variety of
   operating system platforms.  The very large number of systems which
   support this form of file sharing make an SMB URI scheme both
   practical and desirable.

   1.1. Purpose

   This document does not attempt to detail the proper implementation of
   SMB/CIFS or the workings of the NBT transport layer.  The goal is to
   present the syntax of the SMB URI and to describe how that syntax is
   mapped to the semantics of the CIFS suite.  A limited set of
   implementation guidelines is provided in the appendices.

Hertel                     Expires July 8, 2007                 [Page 3]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

2. SMB URI Scheme Overview

   An SMB URI is identified by one of two scheme names: "smb" or "cifs".
   There are existing implementations that support one or the other or
   both of these names, so conforming implementations MUST support both.
   Implementations SHOULD use "smb" as the scheme name when generating
   SMB URIs.

   The SMB URI conforms to the general syntax of URI described in
   [RFC3986].  The scheme is comparable to other URI schemes that are in
   common use, and should be familiar to anyone who uses popular user
   agents such as web browsers.  For example, consider the following


   Both strings might identify the same file, differing only in the
   protocol used to access the file.

   As with other URI schemes, servers (hosts) may be identified by DNS
   name, IPv4 address, or IPv6 address.  The SMB URI scheme also
   supports the use of NetBIOS names to identify SMB services.  NetBIOS
   names are resolved via the NBT Name Service, as described in
   [RFC1001] and detailed in [RFC1002].

3. NetBIOS, NBT, and NBT Transport Semantics

   NBT was created so that applications and services that made use of
   the NetBIOS API could communicate over IP internetworks.  The NBT
   transport maps the semantics of the NetBIOS API onto TCP and UDP,
   hiding the fact that the underlying transport mechanism has changed.

   In order to make this work, RFC1001/1002 define three key services:

   - The NetBIOS Name Service
   - The NetBIOS Datagram Service
   - The NetBIOS Session Service

   These three services collaborate to create a "virtual NetBIOS LAN"
   on top of IP networks.

   The use of NBT as a transport for SMB is widespread, so the syntax
   and semantics of the SMB URI scheme have been adapted to accommodate
   NetBIOS naming conventions and NBT context requirements.

Hertel                     Expires July 8, 2007                 [Page 4]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   3.1. The NetBIOS Name Service

   The NetBIOS API uses names (strings of 16 octets) to identify
   communications endpoints.  These names are the addresses used in
   "NetBIOS LANs".  The NetBIOS Name Service is responsible for
   transparently mapping these names to the IP addresses at which they
   are registered, so that NetBIOS applications and services can find
   one another.

   The set of nodes that are connected to the same virtual NetBIOS LAN
   is known as the "scope" of the LAN.  NBT scopes are always given a
   name, or Scope Identifier (Scope ID), though this normally goes
   unnoticed because the default Scope ID is the empty string ("").
   Multiple scopes, each having a different Scope ID, may intersect
   across the same IP internetwork without conflict.

   The SMB URI scheme provides a mechanism for specifying the NBT
   Scope ID in the URI string.

   3.2. The NetBIOS Datagram Service

   The SMB URI scheme does not make direct use of the Datagram Service,
   so there is no special syntax to support the handling of NetBIOS
   datagram messages.

   3.3. The NetBIOS Session Service

   The Session Service provides a one-to-one mapping of NetBIOS Sessions
   to TCP sessions.  NetBIOS sessions over TCP are initiated by sending
   indicates that the session has been established successfully.  (See
   [RFC1001], section 16.2.)

   The SMB URI scheme supports the exchange of SMB messages over either
   NBT sessions or "naked" TCP sessions.  Other than the SESSION
   REQUEST/POSITIVE SESSION RESPONSE, there is no difference between
   the two transport types.

   3.4  Accommodating NBT in the SMB URI

   The SMB URI scheme supports both NBT and native TCP transport of SMB
   messages.  The syntax of the scheme provides for the inclusion of NBT
   context information so that NetBIOS names can be properly resolved
   and NetBIOS sessions established.

Hertel                     Expires July 8, 2007                 [Page 5]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   The scheme provides the ability to specify the following NBT context

   - Broadcast Address for NBT broadcast name resolution (B mode)
   - NBNS Address for point-to-point name resolution (P mode)
   - Name resolution mode selection (B, P, M, or H mode)
   - NBT Scope ID
   - NetBIOS CALLED name (destination address)
   - NetBIOS CALLING name (source address)

   Most SMB implementations use configuration files or DHCP to
   establish an initial NBT context.  The starting context is referred
   to as the "base context" in the remainder of this document.  NBT
   context information given in absolute URIs is applied against the
   base context to create the "current context".  NBT context
   information given in relative URIs is applied against the current
   context to modify the current context.

   The current NBT context is always used to interpret the meaning of a
   given URI string.  A relative URI containing updated NBT context
   information will cause the resulting URI to be re-evaluated.

4. NetBIOS-Based Workgroups

   The "workgroup" system is a part of the SMB/CIFS protocol suite.
   Workgroups are built on top of NetBIOS, and are identified by their
   NetBIOS names.  The workgroup system allows SMB file servers to be
   organized into named groups, with the goal of making it easier to
   locate resources by categorizing them.

   Within each workgroup, a list of member servers is maintained.  In
   addition to the server list, each workgroup maintains a list of all
   other known workgroups.  The combined list is known as the "browse
   list".  A copy of the browse list may be obtained by sending a
   specific query via SMB.

   The SMB URI scheme views the SMB file sharing environment
   hierarchically.  Conceptually, the hierarchy is arranged as follows:

      List of workgroups (from the browse list)
      + List of servers within a given workgroup (from the browse list)
        + List of shares (shared objects) offered by a server
          + Directories, files, etc. within a share

Hertel                     Expires July 8, 2007                 [Page 6]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   That hierarchy is mapped to the SMB URI scheme as follows:

    URI Format                 Indicates
    =========================  =======================================
      smb://                     List of known workgroups
      smb://smb-wrkgrp/            + SMB servers within the workgroup
      smb://smb-server/              + Shares offered by an SMB server
      smb://smb-server/abs-path        + Directories, files, etc.

   As shown above, the SMB URI provides syntax that indicates requests
   for subsets of the browse list.  In particular, the form:


   represents a request for the list of all known workgroups, while the


   represents a request for the list of servers that are members of a
   particular workgroup.

   A problem arises, however, because the syntax used for requesting the
   list of servers in a workgroup is indistinguishable from that of a
   request for the list of shares (shared objects, such as exported
   filesystems) offered by an SMB file server.  Thus, the two requests
   are differentiated semantically.  Consider the following example:


   If the name "corgi" is a NetBIOS name and it resolves (via the NBT
   Name Service) to a workgroup name, then a user agent would return a
   list of servers in the CORGI workgroup.  If, however, the name
   resolves to a server name, then the user agent would return a list
   of shares offered by the SMB server named CORGI.

   It is rare, but possible (in a misconfigured NBT network), that a
   NetBIOS name will resolve to both a workgroup and an SMB file server.
   In this situation, SMB file services take precedence.  Some user
   agents may be capable of returning both the list of servers in the
   workgroup and the list of shares provided by the SMB file server,
   allowing the user to determine which is the desired result.  This
   is the recommended approach and SHOULD be implemented, where

   There is a special case to be considered when using relative
   references to move between a workgroup reference and a reference to a
   server in the workgroup.  Consider a workgroup named "corgis" and a
   server named "cue" that is a member of that workgroup.

Hertel                     Expires July 8, 2007                 [Page 7]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   Presented with the URI string


   a user agent may return a list of servers that are members of the
   CORGIS workgroup, including node CUE, and allow the user to select
   one of those SMB servers.

   The relative reference used to transition from
   would be "../cue".

   When moving upward in the hierarchy, one might might expect:

      "smb://cue/" + ".." ==> "smb://"

   In this example, however, the user agent is aware that node "cue" is
   a member of the "corgis" workgroup, so:

      "smb://cue/" + ".." ==> "smb://corgis/"

   The NBT workgroup membership of an SMB server may be determined
   either by sending a Node Status Request query to the server (see
   [RFC1001], section 15.1.4) or by maintaining a local cache of
   workgroup information, or both.  Obviously, the choice is
   implementation dependent.  If the server's workgroup membership is
   not available via either of these methods, then it is acceptable to
   move directly to the top of the hierarchy (smb://).

Hertel                     Expires July 8, 2007                 [Page 8]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

5. SMB URI Definition

   The following grammar defines the syntax of the SMB URI.  It is
   based upon the grammar given in Appendix A of [RFC3986].  Refer to
   that RFC for token definitions missing from the grammar below.

      smb-URI        = ( smb-absURI / smb-relURI )
      smb-absURI     = scheme "://" smb-service [ "?" [ nbt-context ] ]
      smb-relURI     = ( path-absolute / path-rootless )
                       [ "?" [ nbt-context ] ]

      scheme         = "smb" / "cifs"
      smb-service    = ( smb-wrkgrp / smb-net-path )

      smb-wrkgrp     = [ smb-userinfo "@" ] [ smb-srv-name ]
                         [ ":" port ] [ "/" ]
      smb-net-path   = smb-server [ path-absolute ]
      smb-server     = [ smb-userinfo "@" ] smb-srv-name [ ":" port ]

      smb-srv-name   = nbt-name / host
      nbt-name       = netbiosname [ "." scope-id ]
      netbiosname    =  netbiosnamec 1*14( netbiosnamec / "*" )
      netbiosnamec   = ( ALPHA / DIGIT / pct-encoded
                         / "-" / "_" / "~"
                         / "!" / "$" / "&" / "'" / "(" / ")"
                         / "+" / "," / ";" / "="

      scope-id       = [ scope-label *( "." scope-label ) ]
      scope-label    = 1*63( scope-char )
      scope-char     = ALPHA / DIGIT / "-" / "_" / "~"
                       / sub-delims / pct-encoded

      smb-userinfo   = [ auth-domain ";" ] userinfo-nosem
      auth-domain    = smb-srv-name
      userinfo-nosem = *( unreserved / pct-encoded
                          / "!" / "$" / "&" / "'" / "(" / ")"
                          / "*" / "+" / "," / "=" / ":"

      nbt-context   = nbt-param *(";" nbt-param )
      nbt-param     = ( "BROADCAST=" IPv4address [ ":" port ]
                        / "CALLED=" netbiosname
                        / "CALLING=" netbiosname
                        / ( "NBNS=" / "WINS=" ) host [ ":" port ]
                        / "NODETYPE=" ("B" / "P" / "M" / "H")
                        / ( "SCOPEID=" / "SCOPE=" ) scope-id

Hertel                     Expires July 8, 2007                 [Page 9]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

6. SMB URI Syntax Elements

   The SMB URI scheme is more or less comparable to other URI schemes
   used for remote filesystem access.  It differs primarily in its
   support for the NBT transport and NBT workgroups.  This section
   provides further explanation and description of those syntax
   elements that are most likely to require clarification.

   6.1. scheme

   As described in section 2, above, an SMB URI is identified by one of
   two scheme names: "smb" or "cifs".  These are equivalent, but the
   former is preferred.

   6.2. smb-service

   The SMB URI can be used to access workgroup information or SMB file
   server services.  There are minor differences in SMB URI syntax
   depending up on which of these service types is being accessed.

   It is possible, for instance, to request workgroup information
   without specifying a destination server name.  In particular,
   the URI:


   represents a request for the list of locally available NBT

   In some situations the workgroup list may not be available to
   unauthenticated users, so the SMB URI scheme allows inclusion of
   smb-userinfo information without the need to specify an smb-srv-name
   (a workgroup name).  Thus, the following is permitted:


   In the above, the username "user" is being supplied.  (The user agent
   should prompt for a password to prevent the password from being

   As with the smb-userinfo field, an SMB URI may include a port
   reference without an smb-srv-name, as in the following example:


   (This is an example of an attempt to retrieve an NBT workgroup list
   via SMB using destination TCP port 4220.)

Hertel                     Expires July 8, 2007                [Page 10]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   Another difference between workgroup and SMB file server references
   is that workgroup references can not be followed by a path.  The
   browse list does not offer shares, directories, or files so an SMB
   URI string such as the following cannot represent a workgroup query:


   6.3. auth-domain

   The auth-domain field is passed to the underlying SMB layer for
   interpretation.  It is used to specify the SMB authentication
   authority (typically a "Domain Controller").  The interpretation of
   this field is specific to the workings of the SMB protocol and should
   be handled by the underlying SMB implementation.

   6.4. smb-srv-name

   The SMB URI supports the use of NetBIOS names and Scope IDs to
   identify SMB servers and services.  When included as part of an SMB
   URI, the syntax of the NetBIOS name is a superset of the syntax of a
   DNS domain name label.  For example:


   Syntactically, the string "jcifs" in the smb-srv-name field of the
   above string may be seen as either a DNS host name (unqualified), or
   as a NetBIOS name.  The underlying SMB implementation MUST determine
   the namespace of the name.  (This is a common problem in SMB
   implementations and is typically solved by first attempting to
   resolve the name as a NetBIOS name and then, if that fails, as a DNS
   host name.)

   Likewise, given:

   the string "" may be interpreted either as a qualified
   DNS name, or as a NetBIOS name with appended Scope ID.

   A NetBIOS name is simply a string of octets with a maximum length of
   15 octets.  (The actual maximum length of the NetBIOS name is
   16-octets, but the 16th is reserved.)  In practice, the only
   restriction on the syntax of a NetBIOS name is that it may not begin
   with an ASCII asterisk character (0x2A).  Octet values that are
   permitted by NetBIOS name syntax but excluded by the SMB URI syntax
   MUST be escaped.  Note, in particular, that the dot character (0x2E)
   MUST be escaped if used in a NetBIOS name.

   The resolution of NetBIOS names to IP addresses is described in
   [RFC1001] and [RFC1002].

Hertel                     Expires July 8, 2007                [Page 11]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   6.5. port

   RFC1001/1002 includes a mechanism for retargeting Session Service
   connections to alternate ports (see [RFC1001], section 16.1.1.)
   which means that non-standard ports may be used for SMB over NBT
   transport.  There may be other valid reasons for providing SMB
   services on on-standard ports.

   The URI port field may be used to specify an alternate service
   port for SMB over either NBT or native TCP transport.  When a
   nonstandard port is used, the transport type (NBT or naked TCP) MUST
   be deduced by the underlying SMB implementation.

   6.6. scope-id

   The correct syntax of an NBT Scope Identifier is described in section
   9 of [RFC1001] as

     "...a character string meeting the requirements of the domain name
     system for domain names."

   The intent was that the Scope ID should match the "preferred syntax"
   for domain names, as given in Appendix 1 of [RFC883] (which has since
   been superceded).  Specifically:

     "The labels must follow the rules for ARPANET host names.  They
     must start with a letter, end with a letter or digit, and have as
     interior characters only letters, digits, and hyphen.  There are
     also some restrictions on the length.  Labels must be 63
     characters or less."

   Several implementations of NBT now exist that disregard the character
   set restrictions described above.  The SMB URI is, therefore,
   a bit more flexible with regard to the octet values that may be
   used.  In fact, the grammar given above allows the use of pct-encoded
   values, so any octet value may be specified.  Note, however, that
   implementations MUST discard zero values ("%00") in the Scope ID
   because many implementations interpret the Scope ID as being a
   nul-terminated string.

   Although the syntax allows Scope IDs that do not match the
   original preferred syntax, the use of nonconforming Scope IDs is
   generally considered unwise because the interpretation of
   nonconforming Scope IDs is likely to be inconsistent.

   The requirement that the Scope ID consist of dot-delimited labels of
   one to sixty-three octets is enforced by the on-the-wire encoding
   used by NBT.

Hertel                     Expires July 8, 2007                [Page 12]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   The SMB URI scheme provides two mechanisms for specifying an NBT
   Scope ID.  The first, as shown in the grammar above, is to append the
   Scope ID to the NetBIOS name as part of the smb-srv-name field, using
   a dot (".") as a delimiter.  This mechanism is included to support
   existing implementations.

   The other mechanism is to specify the Scope ID as part of the
   nbt-context.  The two examples given below are equivalent:


   The latter format is less ambiguous and, therefore, preferred.  User
   agents that rewrite URI strings for display purposes SHOULD rewrite
   SMB URI strings that contain a Scope ID to conform to the nbt-context

   Note also that the scope-id syntax specifically permits a Scope ID
   that is the empty string ("").  The empty string is a valid Scope
   ID and is, in fact, the default on all known implementations.

   6.7. nbt-context

   NBT context information is appended to the tail end of an SMB URI
   string in the form of a URI query string.  Context information is
   specified using key/value pairs.  Multiple context elements may be
   specified by separating the key/value pairs with semi-colons.

   The nbt-context may be used to provide information about the NBT
   transport layer and related support servers.  Information provided
   in the nbt-context overrides the current NBT context maintained by
   the user agent.  The nbt-context is interpreted locally by the user

   The nbt-context is made up of zero or more nbt-params fields, which
   are specified as key/value pairs.  For example:


   In the above example, the CALLED parameter is assigned a value of
   "VIRTSERV", and the NBNS parameter is assigned a value of

   The following keywords are defined:

      BROADCAST:  The IPv4 broadcast address to which to send NBT
                  broadcast name queries.  This may, for example, be
                  used on multi-homed hosts to specify a target subnet.

Hertel                     Expires July 8, 2007                [Page 13]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

                  The value assigned to the BROADCAST keyword may
                  optionally include a port number (delimited by
                  a colon).  The standard port for NBT name resolution
                  is UDP/137.  It is rare that a different port will
                  be used for broadcast name resolution.

         CALLED:  Specifies the NetBIOS name of the SMB server (the
                  NetBIOS destination address.)  A CALLED name is
                  required by the NBT Session Request message (see
                  [RFC1002], Section 4.3.2).

                  If NBT transport is used and the CALLED name is not
                  specified within the URI string, the underlying SMB
                  implementation MUST deduce the CALLED name from
                  available information.  (See Appendix A, below.)

        CALLING:  Specifies the NetBIOS name of the client (the NetBIOS
                  source address.)  This value is only used with NBT
                  transport.  It is required by the NBT Session Request
                  message (see [RFC1002], Section 4.3.2).

                  If NBT transport is used and the CALLING name is not
                  specified in the current NBT context, the underlying
                  SMB implementation MUST generate a suitable name.
                  (Typically, this will be based on the system's host

           NBNS:  Specifies the NetBIOS Name Server (NBNS) to be used
                  for point-to-point NBT Name Resolution.  The NBNS may
                  be specified using a DNS name or an IP address.  See
                  [RFC1001] for information on the NBNS.

                  The value assigned to this parameter may, optionally,
                  include a port number (delimited by a colon).  The
                  standard port for NBT name resolution is UDP/137.
                  The use of a non-standard port for point-to-point NBT
                  name resolution is rare (but less so than it is for
                  broadcast name resolution).

       NODETYPE:  One of B, P, M, H, or the empty string.  These
                  represent the different mechanisms by which a NetBIOS
                  name may be resolved to an IP address on an NBT
                  network.  The first three types are defined in RFC
                  1001/1002.  H mode is the inverse of M mode (in H mode
                  the NBNS is queried before a broadcast query is sent).
                  An empty NODETYPE indicates that NBT name resolution
                  should not be attempted (use DNS name resolution
                  instead).  Some examples:

Hertel                     Expires July 8, 2007                [Page 14]

INTERNET-DRAFT                  SMB URI                  January 8, 2007


          SCOPE:  Specifies the NBT Scope Identifier.  Use of the SCOPE
                  keyword is preferred over inclusion of the Scope ID in
                  the nbt-name field.  User agents MUST support both

                  The default Scope ID is the empty string.  This can be
                  specified in the SMB URI by assigning an empty value
                  to the SCOPE keyword.  For example:


        SCOPEID:  A synonym for SCOPE.

           WINS:  A synonym for NBNS.

   Although all of the keywords and values are shown in upper case, case
   is not significant.

   The client implementation should provide a means for setting the base
   context.  The nbt-context is used to override default values or to
   supply values missing from the local configuration.  Most of all, the
   nbt-context makes it possible for an SMB URI string to maintain a
   consistent interpretation as it travels from one NBT scope to

7. The Relationship Between the SMB URI and the UNC Format

   Some operating systems support a format known as Universal Naming
   Convention (UNC).  UNC is a means for identifying network resources.
   SMB is one of the protocols supported by UNC.

   In general, a UNC string specifying a resource available via SMB
   protocol can be converted into an SMB URI string by simply adding
   the "smb:" (or "cifs:") prefix and reversing the direction of all
   of the separating slashes.  For example:

   UNC form                          URI form
   -------------------------------   -----------------------------------
   \\corgis\docs\                    smb://corgis/docs/
   \\corgis\docs\jolyon\             smb://corgis/docs/jolyon/
   \\corgis\docs\jolyon\rabbit.txt   smb://corgis/docs/jolyon/rabbit.txt

Hertel                     Expires July 8, 2007                [Page 15]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

8. Authentication

   SMB authentication can be divided into the following categories:

      o None
      o Share-based
      o User-based
      o Authentication-Server-based (NT Domain, AD Domain, or Kerberos)

   The authentication mechanism to be used is negotiated during
   client/server session setup.  Client applications, therefore, are
   aware of the server's authentication requirements and may prompt for
   appropriate input (username, password, authentication domain).  By
   prompting for authentication information, an application ensures that
   such information is entered by the user in a controlled manner, and
   that security measures (if any), such as password encryption or
   password hash generation, are applied by the SMB protocol handler
   before the data are transmitted.

   Some authentication values may also be provided within the SMB URI
   string.  In particular, the following fields may optionally be
   included in the URI:

      auth-domain - The authentication domain (single-signon database
                    server) to use for authorization
      userinfo    - User account identifier (username)

9. Security Considerations

   All of the implementations of the SMB URI that are known to exist at
   the time of this writing support inclusion of a password in the
   URI string.  This is generally considered to be dangerous, since
   it may encourage exposure of the plaintext password.  User agents
   should provide a mechanism for prompting the user to enter passwords,
   separate from the URI itself, in a reasonably secure fashion.

   All other general security considerations relate to the workings of
   the SMB/CIFS protocol suite and are beyond the scope of this

Hertel                     Expires July 8, 2007                [Page 16]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

10. Character Encoding Issues

   The only restriction that RFC1001/1002 places on the octet values
   that may be used in a NetBIOS name is that the name may not begin
   with an asterisk ('*', ASCII value 0x2A).  No other values are
   excluded by those RFCs.  For historical reasons, however, some
   implementations disallow the use of a nul byte (0x00) within a
   NetBIOS name.  NetBIOS names are interpreted as fixed-length
   strings of octets, so common mutli-byte character sets may cause
   problems with older implementations.

   Octet values less than 128 (0x80) in a NetBIOS name are typically
   interpreted as US-ASCII characters.  The interpretation of octet
   values above 127 is dependent upon host configuration; there is no
   protocol mechanism to specify which codepage or character set is in
   use.  In general, URI escape sequences should be used to represent
   characters with octet values above 127.

   NetBIOS names, share names, and the directory paths and filenames
   offered by an SMB server may all contain characters from outside the
   7-bit US-ASCII character set.  Applications MUST support the use of
   the URI escape sequence as described in [RFC3986] to accommodate
   octet values that represent non-US-ASCII characters.

   The SMB protocol has evolved over time to include support for various
   character encoding schemes.  A complete discussion of SMB and NBT
   character encoding issues is way beyond the scope of this document.

11. Acknowledgments

   The creation of this document would not have been possible without
   the help and guidance of

   Michael B. Allen
   David Farmer
   Roy T. Fielding
   Steven French
   Larry Masinter
   Richard Sharpe

   and the aggregate knowledge and wisdom of

   The jCIFS Team
   The Samba Team
   The Samba-TNG Team
   The SNIA CIFS Working Group
   The samba-technical and jCIFS mailing list participants
   The IETF URI-review and W3C URI mailing list participants

   Funding for the RFC Editor function is currently provided by the
   Internet Society.

Hertel                     Expires July 8, 2007                [Page 17]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

12. References

     [RFC883] Mockapetris, P., "DOMAIN NAMES - IMPLEMENTATION and
              SPECIFICATION", RFC 883, November 1983.

    [RFC1001] Karl Auerbach, et. al., "Protocol Standard For a NetBIOS
              Service on a TCP/UDP Transport: Concepts and Methods", RFC
              1001, March 1987.

    [RFC1002] Karl Auerbach, et. al., "Protocol Standard For a NetBIOS
              Service on a TCP/UDP Transport: Detailed Specifications",
              RFC 1002, March 1987.

    [RFC3986] Berners-Lee, T., Fielding, R. and L. Masinter, "Uniform
              Resource Identifier (URI): Generic Syntax", RFC 3986,
              January 2005.

   [XOPENSMB] "Protocols for X/Open PC Interworking: SMB, Version 2",
              ISBN 1-872630-45-6, The Open Group, October 1992.

       [ONET] Microsoft Corporation, Intel Corporation, "Microsoft
              Networks/OpenNET Filesharing Protocol", Document Version
              2, Intel Part No. 138446, November 7, 1988.

   [SNIACIFS] Storage Network Industry Association CIFS Documentation
              Work Group, "Common Internet File System (CIFS) Technical
              Reference", Version: CIFS-TR 1.0, March 1, 2002.

    [IMPCIFS] Hertel, Christopher R., "Implementing CIFS -- the Common
              Internet File System", ISBN 0-13-047116-X, Prentice Hall
              PTR, August 2003

13. Author's Address

    Christopher R. Hertel
    Alvarri, Inc.
    15 South Fifth Street
    Suite 1010
    Minneapolis MN 55402


Hertel                     Expires July 8, 2007                [Page 18]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

14. Copyright Notice

    Copyright (C) The Internet Society (2007).  This document is subject
    to the rights, licenses and restrictions contained in BCP 78, and
    except as set forth therein, the authors retain all their rights.

15. Disclaimer of Validity

   This document and the information contained herein are provided on an

Appendix A. Working with NetBIOS Names (Implementation Notes)

   The information presented in this section is intended as a guide for

   Name resolution, particularly with the inclusion of support for
   RFC1001/1002 NBT naming, may result in ambiguous meaning for some SMB
   URI strings.  This problem is reduced if correct NBT context
   information is included in URI strings, and can be eliminated if all
   implementations follow the same basic sequence when resolving server
   names to addresses.

   A.1. NetBIOS Names

   NetBIOS names are addresses.  They represent communication end-points
   within a NetBIOS LAN.  [RFC1001] and [RFC1002] provide a mechanism
   for creating virtual NetBIOS LANs over TCP and UDP transport.  The
   core of that mechanism is the NetBIOS Name Service, which provides
   for mapping between NetBIOS names and the IP addresses at which
   they are registered.  A given host system may register several
   NetBIOS names, each representing an application or service that can
   communicate with other applications or services through the NetBIOS

   A.2. SMB Sessions via NBT

   SMB sessions are established and messages transferred via the NetBIOS
   session service (see [RFC[1001], section 5.3 and [RFC1002] section
   4.3). The system that originates the connection is the "calling"
   node, and the target node is the "called" node.  In order to
   establish an SMB session, a TCP connection is created between the
   calling and called nodes.

Hertel                     Expires July 8, 2007                [Page 19]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   Before a NetBIOS session can be established, the calling node will
   need to discover the IP address of the called node.  This is done
   using the NetBIOS Name Service (see [RFC1001] section 5.2 and
   [RFC1002] section 4.2).  NetBIOS names are always 16 octets, padded
   with spaces (0x20) if necessary, as specified in the RFCs.  By
   convention, however, the 16th octet is reserved for use as a service
   type indicator.  This field is known as the "suffix".

   The suffix byte is NEVER specified in an SMB URI string; it is
   appended by the client implementation.

   A.3. Resolving DNS names and IP addresses to SMB server names

   The NetBIOS Session Service requires that the client provide the
   NetBIOS names of both the calling and called nodes in the NBT
   SESSION REQUEST.  When connecting to an SMB server, the calling name
   (source address) is typically the default NetBIOS name of the client,
   space padded as described, with a suffix byte value of 0x00.  The
   called name (destination address) is the NetBIOS name of the server
   with a suffix byte value of 0x20.

   Applications which support the SMB URI MUST include support for the
   use of DNS names or IP addresses in addition to NetBIOS names when
   initiating SMB connections via NetBIOS over TCP/IP (NBT) transport.
   This functionality is an extension to the NetBIOS over TCP/IP
   behavior specified in RFC 1001 and RFC 1002, and is not part of that
   standard.  It is, however, a common extension and MUST be supported
   for compatibility reasons, and to provide access to SMB shares in
   situations in which the NetBIOS name space cannot be guaranteed to be

   As stated above, the Session Request packet requires a called and a
   calling name, both of which are NetBIOS names.  In order to create an
   NBT Session Request packet, the DNS name or IP address of the server
   MUST be reverse-mapped to the server's NetBIOS name.  Mechanisms for
   doing so include:

   - Issuing a NetBIOS Adapter Status Query

     A NetBIOS Adapter Status Query is sent to the target IP address.
     (See [RFC1001] section 15.1.4 and [RFC1002] sections 4.2.17 &
     4.2.18.)  If a response is received, and if the target node is
     running an SMB server service, then the response will include a
     NetBIOS name with a suffix byte value of 0x20.  This NetBIOS name
     may be used as the called name in a Session Request packet.

     It is possible that multiple entries will have a suffix byte of
     0x20.  If this is the case each name may be tried in turn, or one
     of the other methods MUST be used to discover the name of the SMB
     server service.

Hertel                     Expires July 8, 2007                [Page 20]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   - Generic Server Name

     This method is not supported by all SMB server implementations.

     Some SMB servers will accept the generic SMB server name
     "*SMBSERVER".  A client can simply use the name "*SMBSERVER" as
     the called name in a Session Request packet.  As with all SMB
     server NetBIOS names, the "*SMBSERVER" name MUST be space padded
     and terminated with a suffix byte value of 0x20.

     The "*SMBSERVER" name begins with an asterisk character, so it is
     an illegal NetBIOS name (see [RFC1001], section 5.2).  Thus, it is
     never registered with the NetBIOS Name Service and will not be
     returned in a NetBIOS Adapter Status Response.

     If the target does not support the "*SMBSERVER" generic name, or
     if it is not running SMB services, it will return a CALLED NAME
     NOT PRESENT error.

     Some SMB servers are capable of providing multiple SMB file
     services, each under a different NetBIOS name.  In order to support
     the generic server name, these servers will designate one service
     as a default that will answer to "*SMBSERVER".

   - Parsing the DNS Name or IP address (guessing)

     This is the least reliable method for discovering an SMB server

     Systems which support STD 19 transport will often use the same
     base host name within the DNS and NetBIOS name spaces.  Thus, the
     first label of the DNS name is a good guess at the NetBIOS name of
     the target.  If the input is an IP address rather than a DNS name,
     the a reverse lookup against the DNS may be performed to determine
     the DNS name.

     The first label of the DNS name consists of the initial portion of
     the DNS name string up to but not including the first dot
     character ('.').  If the label is greater than 15 bytes in length,
     it cannot be a NetBIOS name.  The label MUST be space padded to a
     total of 15 bytes, with a suffix value 0x20 (space) added.  This
     forms a valid NetBIOS name that may be used as a called name in a
     Session Request packet.

     If the target returns a CALLED NAME NOT PRESENT error, then the
     DNS name guess is incorrect.

   Any of the above may be tried in any order.

Hertel                     Expires July 8, 2007                [Page 21]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   A.4. Determining the Namespace of the smb-srv-name

   NetBIOS names, DNS names, and IP addresses can not be easily
   distinguished syntactically.  Unfortunately, given the nature and
   history of the SMB/CIFS suite, the appropriate mechanism for
   distinguishing between these server specifier types is the
   trial-and-error method.

   Implementations should begin with the assumption that the specifier
   is a NetBIOS name.  The following process is used to test this

      If the NODETYPE (which is part of the NBT context) is the empty
      string, then no NetBIOS name resolution mechanism has been
      selected and the name cannot be resolved as a NetBIOS name.  Exit.

      If the name string contains dot characters ('.', ASCII 0x2E), then
      separate the name into NetBIOS name and Scope ID at the first dot.
      Otherwise use the entire string as the NetBIOS name, and assume an
      Scope ID of "" unless the Scope ID is specified in the


        If the resulting NetBIOS name is greater than 15 octets in
        length, then the name is not a NetBIOS name.  Exit.

        Issue  STD 19 Name Queries using the NetBIOS name and Scope ID.
        Suffix values of 0x20, 0x1B, and/or 0x1D should be used.  (See
        section A.5., below.)

        If a Positive Name Query Response is received, then the name is
        a NetBIOS name.  Exit, indicating success and returning the
        NetBIOS name and scope ID as parsed.


   If the server specifier is not a NetBIOS name, then it is either a
   DNS name or an IP address.  These are semantically equivalent.

Hertel                     Expires July 8, 2007                [Page 22]

INTERNET-DRAFT                  SMB URI                  January 8, 2007

   A.5. Workgroup vs. SMB Server Names

   If the URI string is of the form


   then the smb-srv-name may represent either an SMB server name or a
   workgroup name.  The name MUST NOT be interpreted as a workgroup name

      - There is path information following the trailing slash.

        Workgroups do not make shares or directories available.

      - The server field is entered as a DNS name or an IP address.

        A workgroup name is a NetBIOS group name.  Workgroups,
        conceptually, represent a group of servers rather than an
        individual server, and the browse list may be retrieved
        from one or more browse servers.

   In these cases, the smb_srv_name is interpreted as a reference to an
   SMB server only.  Workgroups may only be accessed via their NetBIOS

   When testing the name using the algorithm presented in section A.4, a
   NetBIOS name suffix value of 0x20 is used to find an SMB server, and
   a suffix value of 0x1D or 0x1B is used to find a workgroup browse

   A system operating in B mode will use the 0x1D suffix to search for
   a Local Master Browser operating on the same subnet.  A system
   operating in P mode, however, will use the 0x1B suffix to query the
   NBNS for the Domain Master Browser.  An M mode system will first send
   a broadcast query for the 0x1D name and, if that fails, query the
   NBNS for the 0x1B name.  H mode behavior is reverse that of M mode.

Hertel                     Expires July 8, 2007                [Page 23]