RFC 4448






Network Working Group                                    L. Martini, Ed.
Request for Comments: 4448                                      E. Rosen
Category: Standards Track                            Cisco Systems, Inc.
                                                             N. El-Aawar
                                             Level 3 Communications, LLC
                                                                G. Heron
                                                                 Tellabs
                                                              April 2006


   Encapsulation Methods for Transport of Ethernet over MPLS Networks

Status of This Memo



   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

Copyright Notice



   Copyright (C) The Internet Society (2006).

Abstract



   An Ethernet pseudowire (PW) is used to carry Ethernet/802.3 Protocol
   Data Units (PDUs) over an MPLS network.  This enables service
   providers to offer "emulated" Ethernet services over existing MPLS
   networks.  This document specifies the encapsulation of
   Ethernet/802.3 PDUs within a pseudowire.  It also specifies the
   procedures for using a PW to provide a "point-to-point Ethernet"
   service.


















Martini, et al.             Standards Track                     [Page 1]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


Table of Contents



   1. Introduction ....................................................3
   2. Specification of Requirements ...................................6
   3. Applicability Statement .........................................6
   4. Details Specific to Particular Emulated Services ................7
      4.1. Ethernet Tagged Mode .......................................7
      4.2. Ethernet Raw Mode ..........................................8
      4.3. Ethernet-Specific Interface Parameter LDP Sub-TLV ..........8
      4.4. Generic Procedures .........................................9
           4.4.1. Raw Mode vs. Tagged Mode ............................9
           4.4.2. MTU Management on the PE/CE Links ..................11
           4.4.3. Frame Ordering .....................................11
           4.4.4. Frame Error Processing .............................11
           4.4.5. IEEE 802.3x Flow Control Interworking ..............11
      4.5. Management ................................................12
      4.6. The Control Word ..........................................12
      4.7. QoS Considerations ........................................13
   5. Security Considerations ........................................14
   6. PSN MTU Requirements ...........................................14
   7. Normative References ...........................................15
   8. Informative References .........................................15
   9. Significant Contributors .......................................17
   Appendix A. Interoperability Guidelines ...........................20
      A.1. Configuration Options .....................................20
      A.2. IEEE 802.3x Flow Control Considerations ...................21
   Appendix B. QoS Details ...........................................21
      B.1. Adaptation of 802.1Q CoS to PSN CoS .......................22
      B.2. Drop Precedence ...........................................23






















Martini, et al.             Standards Track                     [Page 2]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


1.  Introduction



   An Ethernet pseudowire (PW) allows Ethernet/802.3 [802.3] Protocol
   Data Units (PDUs) to be carried over a Multi-Protocol Label Switched
   [MPLS-ARCH] network.  In addressing the issues associated with
   carrying an Ethernet PDU over a packet switched network (PSN), this
   document assumes that a pseudowire (PW) has been set up by using a
   control protocol such as the one as described in [PWE3-CTRL].  The
   design of Ethernet pseudowire described in this document conforms to
   the pseudowire architecture described in [RFC3985].  It is also
   assumed in the remainder of this document that the reader is familiar
   with RFC 3985.

   The Pseudowire Emulation Edge-to-Edge (PWE3) Ethernet PDU consists of
   the Destination Address, Source Address, Length/Type, MAC Client
   Data, and padding extracted from a MAC frame as a concatenated octet
   sequence in their original order [PDU].

   In addition to the Ethernet PDU format used within the pseudowire,
   this document discusses:

      - Procedures for using a PW in order to provide a pair of Customer
        Edge (CE) routers with an emulated (point-to-point) Ethernet
        service, including the procedures for the processing of Provider
        Edge (PE)-bound and CE-bound Ethernet PDUs [RFC3985]

      - Ethernet-specific quality of service (QoS) and security
        considerations

      - Inter-domain transport considerations for Ethernet PW

   The following two figures describe the reference models that are
   derived from [RFC3985] to support the Ethernet PW emulated services.


















Martini, et al.             Standards Track                     [Page 3]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


            |<-------------- Emulated Service ---------------->|
            |                                                  |
            |          |<------- Pseudowire ------->|          |
            |          |                            |          |
            |          |    |<-- PSN Tunnel -->|    |          |
            | PW End   V    V                  V    V  PW End  |
            V Service  +----+                  +----+  Service V
      +-----+    |     | PE1|==================| PE2|     |    +-----+
      |     |----------|............PW1.............|----------|     |
      | CE1 |    |     |    |                  |    |     |    | CE2 |
      |     |----------|............PW2.............|----------|     |
      +-----+  ^ |     |    |==================|    |     | ^  +-----+
            ^  |       +----+                  +----+     | |  ^
            |  |   Provider Edge 1         Provider Edge 2  |  |
            |  |                                            |  |
      Customer |                                            | Customer
      Edge 1   |                                            | Edge 2
               |                                            |
               |                                            |
      Attachment Circuit (AC)                    Attachment Circuit (AC)
      native Ethernet service                    native Ethernet service

         Figure 1: PWE3 Ethernet/VLAN Interface Reference Configuration

   The "emulated service" shown in Figure 1 is, strictly speaking, a
   bridged LAN; the PEs have MAC interfaces, consume MAC control frames,
   etc.  However, the procedures specified herein only support the case
   in which there are two CEs on the "emulated LAN".  Hence we refer to
   this service as "emulated point-to-point Ethernet".  Specification of
   the procedures for using pseudowires to emulate LANs with more than
   two CEs are out of the scope of the current document.

   +-------------+                                +-------------+
   |  Emulated   |                                |  Emulated   |
   |  Ethernet   |                                |  Ethernet   |
   | (including  |         Emulated Service       | (including  |
   |  VLAN)      |<==============================>|  VLAN)      |
   |  Services   |                                |  Services   |
   +-------------+           Pseudowire           +-------------+
   |Demultiplexer|<==============================>|Demultiplexer|
   +-------------+                                +-------------+
   |    PSN      |            PSN Tunnel          |    PSN      |
   |   MPLS      |<==============================>|   MPLS      |
   +-------------+                                +-------------+
   |  Physical   |                                |  Physical   |
   +-----+-------+                                +-----+-------+

         Figure 2: Ethernet PWE3 Protocol Stack Reference Model



Martini, et al.             Standards Track                     [Page 4]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   For the purpose of this document, PE1 will be defined as the ingress
   router, and PE2 as the egress router.  A layer 2 PDU will be received
   at PE1, encapsulated at PE1, transported, decapsulated at PE2, and
   transmitted out on the attachment circuit of PE2.

   An Ethernet PW emulates a single Ethernet link between exactly two
   endpoints.  The mechanisms described in this document are agnostic to
   that which is beneath the "Pseudowire" level in Figure 2, concerning
   itself only with the "Emulated Service" portion of the stack.

   The following reference model describes the termination point of each
   end of the PW within the PE:

           +-----------------------------------+
           |                PE                 |
   +---+   +-+  +-----+  +------+  +------+  +-+
   |   |   |P|  |     |  |PW ter|  | PSN  |  |P|
   |   |<==|h|<=| NSP |<=|minati|<=|Tunnel|<=|h|<== From PSN
   |   |   |y|  |     |  |on    |  |      |  |y|
   | C |   +-+  +-----+  +------+  +------+  +-+
   | E |   |                                   |
   |   |   +-+  +-----+  +------+  +------+  +-+
   |   |   |P|  |     |  |PW ter|  | PSN  |  |P|
   |   |==>|h|=>| NSP |=>|minati|=>|Tunnel|=>|h|==> To PSN
   |   |   |y|  |     |  |on    |  |      |  |y|
   +---+   +-+  +-----+  +------+  +------+  +-+
           |                                   |
           +-----------------------------------+
                       ^         ^         ^
                       |         |         |
                       A         B         C

           Figure 3: PW Reference Diagram

   The PW terminates at a logical port within the PE, defined at point B
   in the above diagram.  This port provides an Ethernet MAC service
   that will deliver each Ethernet frame that is received at point A,
   unaltered, to the point A in the corresponding PE at the other end of
   the PW.

   The Native Service Processing (NSP) function includes frame
   processing that is required for the Ethernet frames that are
   forwarded to the PW termination point.  Such functions may include
   stripping, overwriting or adding VLAN tags, physical port
   multiplexing and demultiplexing, PW-PW bridging, L2 encapsulation,
   shaping, policing, etc.  These functions are specific to the Ethernet
   technology, and may not be required for the PW emulation service.




Martini, et al.             Standards Track                     [Page 5]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   The points to the left of A, including the physical layer between the
   CE and PE, and any adaptation (NSP) functions between it and the PW
   terminations, are outside of the scope of PWE3 and are not defined
   here.

   "PW Termination", between A and B, represents the operations for
   setting up and maintaining the PW, and for encapsulating and
   decapsulating the Ethernet frames as necessary to transmit them
   across the MPLS network.

   An Ethernet PW operates in one of two modes: "raw mode" or "tagged
   mode".  In tagged mode, each frame MUST contain at least one 802.1Q
   [802.1Q] VLAN tag, and the tag value is meaningful to the NSPs at the
   two PW termination points.  That is, the two PW termination points
   must have some agreement (signaled or manually configured) on how to
   process the tag.  On a raw mode PW, a frame MAY contain an 802.1Q
   VLAN tag, but if it does, the tag is not meaningful to the NSPs, and
   passes transparently through them.

   Additional terminology relevant to pseudowires and Layer 2 Virtual
   Private Networking may be found in [RFC4026].

2.  Specification of Requirements



   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in RFC 2119 [RFC2119].

3.  Applicability Statement



   The Ethernet PW emulation allows a service provider to offer a "port
   to port" Ethernet-based service across an MPLS packet switched
   network (PSN) while the Ethernet VLAN PW emulation allows an
   "Ethernet VLAN to VLAN" based service across an MPLS packet switched
   network (PSN).

   The Ethernet or Ethernet VLAN PW has the following characteristics in
   relationship to the respective native service:

      - An Ethernet PW connects two Ethernet ACs while an Ethernet VLAN
        PW connects two Ethernet VLAN ACs, supporting bidirectional
        transport of variable length Ethernet frames.  The ingress
        Native Service Processing (NSP) function strips the preamble and
        frame check sequence (FCS) from the Ethernet frame and
        transports the frame in its entirety across the PW.  This is
        done regardless of the presence of the 802.1Q tag in the frame.
        The egress NSP function receives the Ethernet frame from the PW
        and regenerates the preamble or FCS before forwarding the frame



Martini, et al.             Standards Track                     [Page 6]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


        to the attachment circuit.  Since the FCS is not transported
        across either Ethernet or Ethernet VLAN PWs, payload integrity
        transparency may be lost.  The OPTIONAL method described in
        [FCS] can be used to achieve payload integrity transparency on
        Ethernet or Ethernet VLAN PWs.

      - For an Ethernet VLAN PW, VLAN tag rewrite can be achieved by NSP
        at the egress PE, which is outside the scope of this document.

      - The Ethernet or Ethernet VLAN PW only supports homogeneous
        Ethernet frame type across the PW; both ends of the PW must be
        either tagged or untagged.  Heterogeneous frame type support
        achieved with NSP functionality is outside the scope of this
        document.

      - Ethernet port or Ethernet VLAN status notification is provided
        using the PW Status TLV in the Label Distribution Protocol (LDP)
        status notification message.  Loss of connectivity between PEs
        can be detected by the LDP session closing, or by using [VCCV]
        mechanisms.  The PE can convey these indications back to its
        attached Remote System.

      - The maximum frame size that can be supported is limited by the
        PSN MTU minus the MPLS header size, unless fragmentation and
        reassembly are used [FRAG].

      - The packet switched network may reorder, duplicate, or silently
        drop packets.  Sequencing MAY be enabled in the Ethernet or
        Ethernet VLAN PW to detect lost, duplicate, or out-of-order
        packets on a per-PW basis.

      - The faithfulness of an Ethernet or Ethernet VLAN PW may be
        increased by leveraging Quality of Service features of the PEs
        and the underlying PSN.  (See Section 4.7, "QoS
        Considerations".)

4.  Details Specific to Particular Emulated Services



4.1.  Ethernet Tagged Mode



   The Ethernet frame will be encapsulated according to the procedures
   defined later in this document for tagged mode.  It should be noted
   that if the VLAN identifier is modified by the egress PE, the
   Ethernet spanning tree protocol might fail to work properly.  If this
   issue is of significance, the VLAN identifier MUST be selected in
   such a way that it matches on the attachment circuits at both ends of
   the PW.




Martini, et al.             Standards Track                     [Page 7]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   If the PE detects a failure on the Ethernet physical port, or the
   port is administratively disabled, it MUST send a PW status
   notification message for all PWs associated with the port.

   This mode uses service-delimiting tags to map input Ethernet frames
   to respective PWs and corresponds to PW type 0x0004 "Ethernet Tagged
   Mode" [IANA].

4.2.  Ethernet Raw Mode



   The Ethernet frame will be encapsulated according to the procedures
   defined later in this document for raw mode.  If the PE detects a
   failure on the Ethernet input port, or the port is administratively
   disabled, the PE MUST send an appropriate PW status notification
   message to the corresponding remote PE.

   In this mode, all Ethernet frames received on the attachment circuit
   of PE1 will be transmitted to PE2 on a single PW.  This service
   corresponds to PW type 0x0005 "Ethernet" [IANA].

4.3.  Ethernet-Specific Interface Parameter LDP Sub-TLV



   This LDP sub-Type Length Value [LDP] specifies interface-specific
   parameters.  When applicable, it MUST be used to validate that the
   PEs, and the ingress and egress ports at the edges of the circuit,
   have the necessary capabilities to interoperate with each other.  The
   Interface parameter TLV is defined in [PWE3-CTRL], the IANA registry
   with initial values for interface parameter sub-TLV types is defined
   in [IANA], but the Ethernet-specific interface parameters are
   specified as follows:

      - 0x06 Requested VLAN ID Sub-TLV

        An Optional 16-bit value indicating the requested VLAN ID.  This
        parameter MUST be used by a PE that is incapable of rewriting
        the 802.1Q Ethernet VLAN tag on output.  If the ingress PE
        receives this request, it MUST rewrite the VLAN ID contained
        inside the VLAN Tag at the input to match the requested VLAN ID.
        If this is not possible, and the VLAN ID does not already match
        the configured ingress VLAN ID, the PW MUST not be enabled.
        This parameter is applicable only to PW type 0x0004.










Martini, et al.             Standards Track                     [Page 8]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


4.4.  Generic Procedures



   When the NSP/Forwarder hands a frame to the PW termination function:

      - The preamble (if any) and FCS are stripped off.

      - The control word as defined in Section 4.6, "The Control Word",
        is, if necessary, prepended to the resulting frame.  The
        conditions under which the control word is or is not used are
        specified below.

      - The proper pseudowire demultiplexer (PW Label) is prepended to
        the resulting packet.

      - The proper tunnel encapsulation is prepended to the resulting
        packet.

      - The packet is transmitted.

   The way in which the proper tunnel encapsulation and pseudowire
   demultiplexer is chosen depends on the procedures that were used to
   set up the pseudowire.

   The tunnel encapsulation depends on how the MPLS PSN is set up.  This
   can include no label, one label, or multiple labels.  The proper
   pseudowire demultiplexer is an MPLS label whose value is determined
   by the PW setup and maintenance protocols.

   When a packet arrives over a PW, the tunnel encapsulation and PW
   demultiplexer are stripped off.  If the control word is present, it
   is processed and stripped off.  The resulting frame is then handed to
   the Forwarder/NSP.  Regeneration of the FCS is considered to be an
   NSP responsibility.

4.4.1.  Raw Mode vs. Tagged Mode



   When the PE receives an Ethernet frame, and the frame has a VLAN tag,
   we can distinguish two cases:

      1. The tag is service-delimiting.  This means that the tag was
         placed on the frame by some piece of service provider-operated
         equipment, and the tag is used by the service provider to
         distinguish the traffic.  For example, LANs from different
         customers might be attached to the same service provider
         switch, which applies VLAN tags to distinguish one customer's
         traffic from another's, and then forwards the frames to the PE.





Martini, et al.             Standards Track                     [Page 9]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


      2. The tag is not service-delimiting.  This means that the tag was
         placed in the frame by a piece of customer equipment, and is
         not meaningful to the PE.

   Whether or not the tag is service-delimiting is determined by local
   configuration on the PE.

   If an Ethernet PW is operating in raw mode, service-delimiting tags
   are NEVER sent over the PW.  If a service-delimiting tag is present
   when the frame is received from the attachment circuit by the PE, it
   MUST be stripped (by the NSP) from the frame before the frame is sent
   to the PW.

   If an Ethernet PW is operating in tagged mode, every frame sent on
   the PW MUST have a service-delimiting VLAN tag.  If the frame as
   received by the PE from the attachment circuit does not have a
   service-delimiting VLAN tag, the PE must prepend the frame with a
   dummy VLAN tag before sending the frame on the PW.  This is the
   default operating mode.  This is the only REQUIRED mode.

   In both modes, non-service-delimiting tags are passed transparently
   across the PW as part of the payload.  It should be noted that a
   single Ethernet packet may contain more than one tag.  At most, one
   of these tags may be service-delimiting.  In any case, the NSP
   function may only inspect the outermost tag for the purpose of
   adapting the Ethernet frame to the pseudowire.

   In both modes, the service-delimiting tag values have only local
   significance, i.e., are meaningful only at a particular PE-CE
   interface.  When tagged mode is used, the PE that receives a frame
   from the PW may rewrite the tag value, or may strip the tag entirely,
   or may leave the tag unchanged, depending on its configuration.  When
   raw mode is used, the PE that receives a frame may or may not need to
   add a service-delimiting tag before transmitting the frame on the
   attachment circuit; however, it MUST not rewrite or remove any tags
   that are already present.

   The following table illustrates the operations that might be
   performed at input from the attachment circuit:

   +-----------------------------------------------------------+
   |       Tag-> |  service delimiting | non service delimiting|
   |-------------+---------------------+-----------------------|
   |   Raw Mode  | 1st VLAN Tag Removed| no operation performed|
   |-------------+---------------------+-----------------------|
   | Tagged Mode | NO OP or Tag Added  |     Tag Added         |
   +-----------------------------------------------------------+




Martini, et al.             Standards Track                    [Page 10]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


4.4.2.  MTU Management on the PE/CE Links



   The Ethernet PW MUST NOT be enabled unless it is known that the MTUs
   of the CE-PE links are the same at both ends of the PW.  If an egress
   router receives an encapsulated layer 2 PDU whose payload length
   (i.e., the length of the PDU itself without any of the encapsulation
   headers) exceeds the MTU of the destination layer 2 interface, the
   PDU MUST be dropped.

4.4.3.  Frame Ordering



   In general, applications running over Ethernet do not require strict
   frame ordering.  However, the IEEE definition of 802.3 [802.3]
   requires that frames from the same conversation in the context of
   link aggregation (clause 43) are delivered in sequence.  Moreover,
   the PSN cannot (in the general case) be assumed to provide or to
   guarantee frame ordering.  An Ethernet PW can, through use of the
   control word, provide strict frame ordering.  If this option is
   enabled, any frames that get misordered by the PSN will be dropped or
   reordered by the receiving PW endpoint.  If strict frame ordering is
   a requirement for a particular PW, this option MUST be enabled.

4.4.4.  Frame Error Processing



   An encapsulated Ethernet frame traversing a pseudowire may be
   dropped, corrupted, or delivered out-of-order.  As described in
   [PWE3-REQ], frame loss, corruption, and out-of-order delivery are
   considered to be a "generalized bit error" of the pseudowire.  PW
   frames that are corrupted will be detected at the PSN layer and
   dropped.

   At the ingress of the PW, the native Ethernet frame error processing
   mechanisms MUST be enabled.  Therefore, if a PE device receives an
   Ethernet frame containing hardware-level Cyclic Redundancy Check
   (CRC) errors, framing errors, or a runt condition, the frame MUST be
   discarded on input.  Note that defining this processing is part of
   the NSP function and is outside the scope of this document.

4.4.5.  IEEE 802.3x Flow Control Interworking



   In a standard Ethernet network, the flow control mechanism is
   optional and typically configured between the two nodes on a point-
   to-point link (e.g., between the CE and the PE).  IEEE 802.3x PAUSE
   frames MUST NOT be carried across the PW.  See Appendix A for notes
   on CE-PE flow control.






Martini, et al.             Standards Track                    [Page 11]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


4.5.  Management



   The Ethernet PW management model follows the general PW management
   model defined in [RFC3985] and [PWE3-MIB].  Many common PW management
   facilities are provided here, with no additional Ethernet specifics
   necessary.  Ethernet-specific parameters are defined in an additional
   MIB module, [PW-MIB].

4.6.  The Control Word



   The control word defined in this section is based on the Generic PW
   MPLS Control Word as defined in [PWE3-CW].  It provides the ability
   to sequence individual frames on the PW, avoidance of equal-cost
   multiple-path load-balancing (ECMP) [RFC2992], and Operations and
   Management (OAM) mechanisms including VCCV [VCCV].

   [PWE3-CW] states, "If a PW is sensitive to packet misordering and is
   being carried over an MPLS PSN that uses the contents of the MPLS
   payload to select the ECMP path, it MUST employ a mechanism which
   prevents packet misordering." This is necessary because ECMP
   implementations may examine the first nibble after the MPLS label
   stack to determine whether the labelled packet is IP or not.  Thus,
   if the source MAC address of an Ethernet frame carried over the PW
   without a control word present begins with 0x4 or 0x6, it could be
   mistaken for an IPv4 or IPv6 packet.  This could, depending on the
   configuration and topology of the MPLS network, lead to a situation
   where all packets for a given PW do not follow the same path.  This
   may increase out-of-order frames on a given PW, or cause OAM packets
   to follow a different path than actual traffic (see Section 4.4.3,
   "Frame Ordering").

   The features that the control word provides may not be needed for a
   given Ethernet PW.  For example, ECMP may not be present or active on
   a given MPLS network, strict frame sequencing may not be required,
   etc.  If this is the case, the control word provides little value and
   is therefore optional.  Early Ethernet PW implementations have been
   deployed that do not include a control word or the ability to process
   one if present.  To aid in backwards compatibility, future
   implementations MUST be able to send and receive frames without the
   control word present.

   In all cases, the egress PE MUST be aware of whether the ingress PE
   will send a control word over a specific PW.  This may be achieved by
   configuration of the PEs, or by signaling, as defined in [PWE3-CTRL].







Martini, et al.             Standards Track                    [Page 12]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   The control word is defined as follows:

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |0 0 0 0|   Reserved            |       Sequence Number         |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   In the above diagram, the first 4 bits MUST be set to 0 to indicate
   PW data.  The rest of the first 16 bits are reserved for future use.
   They MUST be set to 0 when transmitting, and MUST be ignored upon
   receipt.

   The next 16 bits provide a sequence number that can be used to
   guarantee ordered frame delivery.  The processing of the sequence
   number field is OPTIONAL.

   The sequence number space is a 16-bit, unsigned circular space.  The
   sequence number value 0 is used to indicate that the sequence number
   check algorithm is not used.  The sequence number processing
   algorithm is found in [PWE3-CW].

4.7.  QoS Considerations



   The ingress PE MAY consider the user priority (PRI) field [802.1Q] of
   the VLAN tag header when determining the value to be placed in a QoS
   field of the encapsulating protocol (e.g., the EXP fields of the MPLS
   label stack).  In a similar way, the egress PE MAY consider the QoS
   field of the encapsulating protocol (e.g., the EXP fields of the MPLS
   label stack) when queuing the frame for transmission towards the CE.

   A PE MUST support the ability to carry the Ethernet PW as a best-
   effort service over the MPLS PSN.  PRI bits are kept transparent
   between PE devices, regardless of the QoS support of the PSN.

   If an 802.1Q VLAN field is added at the PE, a default PRI setting of
   zero MUST be supported, a configured default value is recommended, or
   the value may be mapped from the QoS field of the PSN, as referred to
   above.

   A PE may support additional QoS support by means of one or more of
   the following methods:

        i.  One class of service (CoS) per PW End Service (PWES), mapped
            to a single CoS PW at the PSN.
       ii.  Multiple CoS per PWES mapped to a single PW with multiple
            CoS at the PSN.
      iii.  Multiple CoS per PWES mapped to multiple PWs at the PSN.



Martini, et al.             Standards Track                    [Page 13]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   Examples of the cases above and details of the service mapping
   considerations are described in Appendix B.

   The PW guaranteed rate at the MPLS PSN level is PW service provider
   policy based on agreement with the customer, and may be different
   from the Ethernet physical port rate.

5.  Security Considerations



   The Ethernet pseudowire type is subject to all of the general
   security considerations discussed in [RFC3985] and [PWE3-CTRL].

   The Ethernet pseudowire is transported on an MPLS PSN; therefore, the
   security of the pseudowire itself will only be as good as the
   security of the MPLS PSN.  The MPLS PSN can be secured by various
   methods, as described in [MPLS-ARCH].

   Security achieved by access control of MAC addresses is out of the
   scope of this document.  Additional security requirements related to
   the use of PW in a switching (virtual bridging) environment are not
   discussed here as they are not within the scope of this document.

6.  PSN MTU Requirements



   The MPLS PSN MUST be configured with an MTU that is large enough to
   transport a maximum-sized Ethernet frame that has been encapsulated
   with a control word, a pseudowire demultiplexer, and a tunnel
   encapsulation.  With MPLS used as the tunneling protocol, for
   example, this is likely to be 8 or more bytes greater than the
   largest frame size.  The methodology described in [FRAG] MAY be used
   to fragment encapsulated frames that exceed the PSN MTU.  However, if
   [FRAG] is not used and if the ingress router determines that an
   encapsulated layer 2 PDU exceeds the MTU of the PSN tunnel through
   which it must be sent, the PDU MUST be dropped.

















Martini, et al.             Standards Track                    [Page 14]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


7.  Normative References



   [PWE3-CW]    Bryant, S., Swallow, G., and D. McPherson, "Pseudowire
                Emulation Edge-to-Edge (PWE3) Control Word for Use over
                an MPLS PSN", RFC 4385, February 2006.

   [IANA]       Martini, L., "IANA Allocations for Pseudowire Edge to
                Edge Emulation (PWE3)", BCP 116, RFC 4446, April 2006.

   [PWE3-CTRL]  Martini, L., El-Aawar, N., Heron, G., Rosen, E., Tappan,
                D., and T. Smith, "Pseudowire Setup and Maintenance
                using the Label Distribution Protocol (LDP)", RFC 4447,
                April 2006.

   [MPLS-ARCH]  Rosen, E., Viswanathan, A., and R. Callon,
                "Multiprotocol Label Switching Architecture", RFC 3031,
                January 2001.

   [802.3]      IEEE802.3-2005, ISO/IEC 8802-3: 2000 (E), "IEEE Standard
                for Information technology -- Telecommunications and
                information exchange between systems -- Local and
                metropolitan
                 area networks -- Specific requirements -- Part 3:
                Carrier Sense Multiple Access with Collision Detection
                (CSMA/CD) Access Method and Physical Layer
                Specifications", 2005.

   [802.1Q]     ANSI/IEEE Standard 802.1Q-2005, "IEEE Standards for
                Local and Metropolitan Area Networks: Virtual Bridged
                Local Area Networks", 2005.

   [PDU]        IEEE Std 802.3, 1998 Edition, "Part 3: Carrier sense
                multiple access with collision detection (CSMA/CD)
                access method and physical layer specifications" figure
                3.1, 1998

   [RFC2119]    Bradner, S., "Key words for use in RFCs to Indicate
                Requirement Levels", BCP 14, RFC 2119, March 1997.

8.  Informative References



   [RFC3985]    Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
                Edge (PWE3) Architecture", RFC 3985, March 2005.

   [PW-MIB]     Zelig, D. and T. Nadeau, "Ethernet Pseudo Wire (PW)
                Management Information Base", Work in Progress, February
                2006.




Martini, et al.             Standards Track                    [Page 15]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   [PWE3-REQ]   Xiao, X., McPherson, D., and P. Pate, "Requirements for
                Pseudo-Wire Emulation Edge-to-Edge (PWE3)", RFC 3916,
                September 2004.

   [PWE3-MIB]   Zelig, D., Ed. and T. Nadeau, Ed., "Pseudo Wire (PW)
                Management Information Base", Work in Progress, February
                2006.

   [LDP]        Andersson, L., Doolan, P., Feldman, N., Fredette, A.,
                and B. Thomas, "LDP Specification", RFC 3036, January
                2001.

   [FRAG]       Malis, A. and W. Townsley, "PWE3 Fragmentation and
                Reassembly", Work in Progress, February 2005.

   [FCS]        Malis, A., Allan, D., and N. Del Regno, "PWE3 Frame
                Check Sequence Retention", Work in Progress, September
                2005.

   [VCCV]       Nadeau, T., Ed. and R. Aggarwal, Ed., "Pseudo Wire
                Virtual Circuit Connectivity Verification (VCCV)", Work
                in Progress, August 2005.

   [RFC2992]    Hopps, C., "Analysis of an Equal-Cost Multi-Path
                Algorithm", RFC 2992, November 2000.

   [RFC4026]    Andersson, L. and T. Madsen, "Provider Provisioned
                Virtual Private Network (VPN) Terminology", RFC 4026,
                March 2005.

   [L2TPv3]     Lau, J., Townsley, M., and I. Goyret, "Layer Two
                Tunneling Protocol - Version 3 (L2TPv3)", RFC 3931,
                March 2005.


















Martini, et al.             Standards Track                    [Page 16]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


9.  Significant Contributors



   Andrew G. Malis
   Tellabs
   90 Rio Robles Dr.
   San Jose, CA 95134

   EMail: Andy.Malis@tellabs.com


   Dan Tappan
   Cisco Systems, Inc.
   1414 Massachusetts Avenue
   Boxborough, MA 01719

   EMail: tappan@cisco.com


   Steve Vogelsang
   ECI Telecom
   Omega Corporate Center
   1300 Omega Drive
   Pittsburgh, PA 15205

   EMail: stephen.vogelsang@ecitele.com


   Vinai Sirkay
   Reliance Infocomm
   Dhirubai Ambani Knowledge City
   Navi Mumbai 400 709
   India

   EMail: vinai@sirkay.com


   Vasile Radoaca
   Nortel Networks
   600  Technology Park
   Billerica MA 01821

   EMail: vasile@nortelnetworks.com









Martini, et al.             Standards Track                    [Page 17]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   Chris Liljenstolpe
   Alcatel
   11600 Sallie Mae Dr.
   9th Floor
   Reston, VA 20193

   EMail: chris.liljenstolpe@alcatel.com


   Kireeti Kompella
   Juniper Networks
   1194 N. Mathilda Ave
   Sunnyvale, CA 94089

   EMail: kireeti@juniper.net


   Tricci So
   Nortel Networks 3500 Carling Ave.,
   Nepean, Ontario,
   Canada, K2H 8E9.

   EMail: tso@nortelnetworks.com


   XiPeng Xiao
   Riverstone Networks
   5200 Great America Parkway
   Santa Clara, CA 95054

   EMail: xxiao@riverstonenet.com


   Christopher O.  Flores
   T-Systems
   10700 Parkridge Boulevard
   Reston, VA 20191
   USA

   EMail: christopher.flores@usa.telekom.de


   David Zelig
   Corrigent Systems
   126, Yigal Alon St.
   Tel Aviv, ISRAEL

   EMail: davidz@corrigent.com



Martini, et al.             Standards Track                    [Page 18]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


   Raj Sharma
   Luminous Networks, Inc.
   10460 Bubb Road
   Cupertino, CA 95014

   EMail: raj@luminous.com


   Nick Tingle
   TiMetra Networks
   274 Ferguson Drive
   Mountain View, CA 94043

   EMail: nick@timetra.com


   Sunil Khandekar
   TiMetra Networks
   274 Ferguson Drive
   Mountain View, CA 94043

   EMail: sunil@timetra.com


   Loa Andersson
   TLA-group

   EMail: loa@pi.se























Martini, et al.             Standards Track                    [Page 19]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


Appendix A.  Interoperability Guidelines



A.1.  Configuration Options



   The following is a list of the configuration options for a point-to-
   point Ethernet PW based on the reference points of Figure 3:

   --------------|---------------|---------------|------------------
   Service and   |  Encap on C   |Operation at B | Remarks
   Encap on A    |               |ingress/egress |
   --------------|---------------|---------------|------------------
   1) Raw        | Raw - Same as |               |
                 | A             |               |
                 |               |               |
   --------------|---------------|---------------|------------------
   2) Tag1       | Tag2          |Optional change| VLAN can be
                 |               |of VLAN value  | 0-4095
                 |               |               | Change allowed in
                 |               |               | both directions
   --------------|---------------|---------------|------------------
   3) No Tag     | Tag           |Add/remove Tag | Tag can be
                 |               |field          | 0-4095
                 |               |               | (note i)
                 |               |               |
   --------------|---------------|---------------|------------------
   4) Tag        | No Tag        |Remove/add Tag | (note ii)
                 |               |field          |
                 |               |               |
                 |               |               |
   --------------|---------------|---------------|------------------

                      Figure 4: Configuration Options

   Allowed combinations:

   Raw and other services are not allowed on the same NSP virtual port
   (A).  All other combinations are allowed, except that conflicting
   VLANs on (A) are not allowed.  Note that in most point-to-point PW
   applications the NSP virtual port is the same entity as the physical
   port.

   Notes:

        i.  Mode #3 MAY be limited to adding VLAN NULL only, since
            change of VLAN or association to specific VLAN can be done
            at the PW CE-bound side.





Martini, et al.             Standards Track                    [Page 20]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


       ii.  Mode #4 exists in layer 2 switches, but is not recommended
            when operating with PW since it may not preserve the user's
            PRI bits.  If there is a need to remove the VLAN tag (for
            TLS at the other end of the PW), it is recommended to use
            mode #2 with tag2=0 (NULL VLAN) on the PW and use mode #3 at
            the other end of the PW.

A.2.  IEEE 802.3x Flow Control Considerations



   If the receiving node becomes congested, it can send a special frame,
   called the PAUSE frame, to the source node at the opposite end of the
   connection.  The implementation MUST provide a mechanism for
   terminating PAUSE frames locally (i.e., at the local PE).  It MUST
   operate as follows: PAUSE frames received on a local Ethernet port
   SHOULD cause the PE device to buffer, or to discard, further Ethernet
   frames for that port until the PAUSE condition is cleared.
   Optionally, the PE MAY simply discard PAUSE frames.

   If the PE device wishes to pause data received on a local Ethernet
   port (perhaps because its own buffers are filling up or because it
   has received notification of congestion within the PSN), then it MAY
   issue a PAUSE frame on the local Ethernet port, but MUST clear this
   condition when willing to receive more data.

Appendix B.  QoS Details



   Section 4.7, "QoS Considerations", describes various modes for
   supporting PW QOS over the PSN.  Examples of the above for a point-
   to-point VLAN service are:

      - The classification to the PW is based on VLAN field, but the
        user PRI bits are mapped to different CoS markings (and network
        behavior) at the PW level.  An example of this is a PW mapped to
        an E-LSP in an MPLS network.

      - The classification to the PW is based on VLAN field and the PRI
        bits, and frames with different PRI bits are mapped to different
        PWs.  An example is to map a PWES to different L-LSPs in MPLS
        PSN in order to support multiple CoS over an L-LSP-capable
        network, or to map a PWES to multiple L2TPv3 sessions [L2TPv3].

        The specific value to be assigned at the PSN for various CoS is
        out of the scope of this document.








Martini, et al.             Standards Track                    [Page 21]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


B.1.  Adaptation of 802.1Q CoS to PSN CoS



   It is not required that the PSN will have the same CoS definition of
   CoS as defined in [802.1Q], and the mapping of 802.1Q CoS to PSN CoS
   is application specific and depends on the agreement between the
   customer and the PW provider.  However, the following principles
   adopted from 802.1Q, Table 8-2, MUST be met when applying the set of
   PSN CoS based on user's PRI bits.

                ----------------------------------
                |#of available classes of service|
   -------------||---+---+---+---+---+---+---+---|
   User         || 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
   Priority     ||   |   |   |   |   |   |   |   |
   ===============================================
   0 Best Effort|| 0 | 0 | 0 | 1 | 1 | 1 | 1 | 2 |
   (Default)    ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|
   1 Background || 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
                ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|
   2 Spare      || 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 |
                ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|
   3 Excellent  || 0 | 0 | 0 | 1 | 1 | 2 | 2 | 3 |
   Effort       ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|
   4 Controlled || 0 | 1 | 1 | 2 | 2 | 3 | 3 | 4 |
   Load         ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|
   5 Interactive|| 0 | 1 | 1 | 2 | 3 | 4 | 4 | 5 |
   Multimedia   ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|
   6 Interactive|| 0 | 1 | 2 | 3 | 4 | 5 | 5 | 6 |
   Voice        ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|
   7 Network    || 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
   Control      ||   |   |   |   |   |   |   |   |
   ------------ ||---+---+---+---+---+---+---+---|

                     Figure 5: IEEE 802.1Q CoS Mapping










Martini, et al.             Standards Track                    [Page 22]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


B.2.  Drop Precedence



   The 802.1P standard does not support drop precedence; therefore, from
   the PW PE-bound point of view there is no mapping required.  It is,
   however, possible to mark different drop precedence for different PW
   frames based on the operator policy and required network behavior.
   This functionality is not discussed further here.

   PSN QoS support and signaling of QoS are out of the scope of this
   document.

Authors' Addresses



   Luca Martini, Editor
   Cisco Systems, Inc.
   9155 East Nichols Avenue, Suite 400
   Englewood, CO, 80112

   EMail: lmartini@cisco.com


   Nasser El-Aawar
   Level 3 Communications, LLC.
   1025 Eldorado Blvd.
   Broomfield, CO, 80021

   EMail: nna@level3.net


   Giles Heron
   Tellabs
   Abbey Place
   24-28 Easton Street
   High Wycombe
   Bucks
   HP11 1NT
   UK

   EMail: giles.heron@tellabs.com


   Eric C. Rosen
   Cisco Systems, Inc.
   1414 Massachusetts Avenue
   Boxborough, MA 01719

   EMail: erosen@cisco.com




Martini, et al.             Standards Track                    [Page 23]

RFC 4448          Encapsulation of Ethernet over MPLS         April 2006


Full Copyright Statement



   Copyright (C) The Internet Society (2006).

   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.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET
   ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED,
   INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE
   INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property



   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at
   ietf-ipr@ietf.org.

Acknowledgement



   Funding for the RFC Editor function is provided by the IETF
   Administrative Support Activity (IASA).







Martini, et al.             Standards Track                    [Page 24]