Network Working Group Luca Martini Internet Draft Nasser El-Aawar Expiration Date: November 2000 Level 3 Communications, LLC. Dimitri Stratton Vlachos Daniel Tappan Eric C. Rosen Cisco Systems, Inc. May 2000 Transport of Layer 2 Frames Over MPLS draft-martini-l2circuit-trans-mpls-01.txt Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. 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 http://www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. Abstract This document described a method for transporting the Protocol Data Units (PDUs) of layer 2 protocols such as Frame Relay, ATM AAL5, and ethernet across an MPLS network. Martini, et al. [Page 1] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 Table of Contents 1 Specification of Requirements .......................... 2 2 Introduction ........................................... 2 3 Tunnel Labels and VC Labels ............................ 3 4 Optional Sequencing and/or Padding ..................... 4 5 Protocol-Specific Issues ............................... 5 5.1 Frame Relay ............................................ 5 5.2 ATM .................................................... 5 5.2.1 F5 OAM Cell Support .................................... 6 5.2.2 CLP Bit ................................................ 6 5.2.3 PTI Field in ATM Cell Mode ............................. 6 5.3 Ethernet VLAN .......................................... 7 6 LDP .................................................... 7 7 Security Considerations ................................ 9 8 References ............................................. 9 9 Author Information ..................................... 10 1. 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. 2. Introduction In an MPLS network, it is possible to carry the Protocol Data Units (PDUs) of layer 2 protocols by prepending an MPLS label stack to these PDUs. This document specifies the necessary label distribution and encapsulation procedures for accomplishing this. We restrict discussion to the case of point-to-point transport. QoS related issues are not discussed in this draft. Martini, et al. [Page 2] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 3. Tunnel Labels and VC Labels Suppose it is desired to transport layer 2 PDUs from ingress LSR R1 to egress LSR R2, across an intervening MPLS network. We assume that there is an LSP from R1 to R2. That is, we assume that R1 can cause a packet to be delivered to R2 by pushing some label onto the packet and sending the result to one of its adjacencies. Call this label the "tunnel label", and the corresponding LSP the "tunnel LSP". The tunnel LSP merely gets packets from R1 to R2, the corresponding label doesn't tell R2 what to do with the payload, and in fact if penultimate hop popping is used, R2 may never even see the corresponding label. (If R1 itself is the penultimate hop, a tunnel label may not even get pushed on.) Thus if the payload is not an IP packet, there must be a label, which becomes visible to R2, that tells R2 how to treat the received packet. Call this label the "VC label". So when R1 sends a layer 2 PDU to R2, it first pushes a VC label on its label stack, and then (if R1 is not adjacent to R2) pushes on a tunnel label. The tunnel label gets the MPLS packet from R1 to R2; the VC label is not visible until the MPLS packet reaches R2. R2's disposition of the packet is based on the VC label. If the payload of the MPLS packet is, for example, an ATM AAL5 PDU, the VC label will generally correspond to a particular ATM VC at R2. That is, R2 needs to be able to infer from the VC label the outgoing interface and the VPI/VCI value for the AAL5 PDU. If the payload is a Frame Relay PDU, then R2 needs to be able to infer from the VC label the outgoing interface and the DLCI value. If the payload is an ethernet frame, then R2 needs to be able to infer from the VC label the outgoing interface, and perhaps the VLAN identifier. Note that the VC label must always be at the bottom of the label stack, and the tunnel label, if present, must be immediately above the VC label. Of course, as the packet is transported across the MPLS network, additional labels may be pushed on (and then popped off) as needed. Even R1 itself may push on additional labels above the tunnel label. If R1 and R2 are directly adjacent LSRs, then it may not be necessary to use a tunnel label at all. This document does not specify a method for distributing the tunnel label or any other labels that may appear above it on the stack. Any acceptable method of MPLS label distribution will do. This document does specify a method for assigning and distributing the VC label. Static label assignment MAY be used, and implementations SHOULD provide support for this. If signalling is Martini, et al. [Page 3] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 used, the VC label MUST be distributed from R2 to R1 using LDP in the downstream unsolicited mode; this requires that an LDP connection be created between R1 and R2. Note that this technique allows an unbounded number of layer 2 "VCs" to be carried together in a single "tunnel". Thus it scales quite well in the network backbone. 4. Optional Sequencing and/or Padding Sometimes it is important to guarantee that sequentiality is preserved on a layer 2 virtual circuit. To accommodate this requirement, we provide an optional control word which may appear immediately after the label stack and immediately before the layer 2 PDU. This control word contains a sequence number. R1 and R2 both need to be configured with the knowledge of whether a control word will be used for a specific virtual circuit. Sometimes it is necessary to transmit a small packet on a medium where there is a minimum transport unit larger than the actual packet size. In this case, padding is appended to the packet. When the VC label is popped, it may be desirable to remove the padding before forwarding the packet. To facilitate this, the control word has a length field. If the packet's length (without any padding) is less than 256 bytes, the length field MUST be set to the packet's length (without padding). Otherwise the length field MUST be set to zero. The value of the length field, if non-zero, can be used to remove any padding. 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 2 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Reserved | Length | Sequence Number | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ The first 8 bits are reserved for future use. They MUST be set to 0 when transmitting, and MUST be ignored upon receipt. The length byte is set as specified above. The next 16 bits are the sequence number that is used to guarantee ordered packet delivery. For a given VC label, and a given pair of LSRs, R1 and R2, where R2 has distributed that VC label to R1, the sequence number is initialized to 0, and is incremented by one for Martini, et al. [Page 4] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 each successive packet carrying that VC label which R1 transmits to R2. The sequence number space is a 16 bit unsigned circular space. PDUs carrying the control word MUST NOT be delivered out of order. They may be discarded or reordered. 5. Protocol-Specific Issues 5.1. Frame Relay A Frame Relay PDU is transported in its entirety, including the Frame Relay Header. The sequencing control word is optional. The BCN and FCN signals are carried unchanged across the network in the frame relay header. These signals do not appear in the MPLS header, and are unseen by the MPLS network. If the MPLS edge LSR detects a service affecting condition as defined in [2] Q.933 Annex A.5 sited in IA FRF1.1, it will withdraw the label that corresponds to the frame relay DLCI. The Egress side should generate the corresponding errors and alarms as defined in [2] on the Frame relay VC. The ingress LSR MAY consider the DE bit of the Frame Relay header when determining the value to be placed in the EXP fields of the MPLS label stack. In a similar way, the egress LSR MAY consider the EXP field of the VC label when queuing the packet for egress. 5.2. ATM Two modes are supported for ATM transport, ATM Adaptation Layer 5 (AAL5) and ATM cell. In ATM AAL5 mode the ingress LSR is required to reassemble AAL5 CPCS-PDUs from the incoming VC and transport each CPCS-PDU as a single packet. No AAL5 trailer is transported. The sequencing control word is optional. In ATM cell mode the ingress LSR transports each ATM cell payload as a single packet. No ATM cell header is transported. The sequencing control word is OPTIONAL. Martini, et al. [Page 5] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 5.2.1. F5 OAM Cell Support F5 OAM cells are not transported on the VC LSP. If an F5 end-to-end OAM cell is received from a VC by a LSR with a loopback indication value of 1 and the LSR has a label mapping for the VC, the LSR must decrement the loopback indication value and loop back the cell on the VC. Otherwise the loopback cell must be silently discarded by the LSR. A LSR may optionally be configured to periodically generate F5 end- to-end loopback OAM cells on a VC. In this case, the LSR must only generate F5 end-to-end loopback cells while a label mapping exists for the VC. If the VC label mapping is withdrawn the LSR must cease generation of F5 end-to-end loopback OAM cells. If the LSR fails to receive a response to an F5 end-to-end loopback OAM cell for a pre- defined period of time it must withdraw the label mapping for the VC. If an ingress LSR receives an AIS F5 OAM cell, fails to receive a pre-defined number of the End-to-End loop OAM cells, or a physical interface goes down, it must withdraw the label mappings for all VCs associated with the failure. When a VC label mapping is withdrawn, the egress LSR must generate AIS F5 OAM cells on the VC associated with the withdrawn label mapping. 5.2.2. CLP Bit The ingress LSR MAY consider the CLP bit when determining the value to be placed in the EXP fields of the MPLS label stack. The egress LSR MAY consider the value of the EXP field of the VC label when determining the value of the ATM CLP bit. 5.2.3. PTI Field in ATM Cell Mode ATM cell mode is intended for transporting non-AAL5 traffic only. The ingress LSR must transport cells with a PTI of 0. Cells with a PTI other than 0 are not transported on the LSP. The egress LSR must set the PTI to 0 for cells switched from a VC LSP to an outgoing VC. Martini, et al. [Page 6] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 5.3. Ethernet VLAN For and ethernet 802.1q VLAN the entire ethernet frame without the preamble is transported as a single packet. The sequencing control word is OPTIONAL. If a packet is received out of sequence it MUST be dropped. The VLAN 4 byte tag is transported as is, and MAY be overwritten by the egress LSR. The ingress LSR MAY consider the user priority field [4] of the VLAN tag header when determining the value to be placed in the EXP fields of the MPLS label stack. In a similar way, the egress LSR MAY consider the EXP field of the VC label when queuing the packet for egress. 6. LDP The VC label bindings are distributed using the LDP downstream unsolicited mode described in [1]. The LSRs will establish and LDP session using the Extended Discovery mechanism described in [1, section 2.4-2.5] A new type of FEC TLV, a VC FEC Element type 128 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 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | VC tlv | VC Type | VC ID len | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Group ID | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | VC ID | | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ - VC Type A two octet quantity containing a value which represents the type of VC. Assigned Values are: 1 - Frame Relay DLCI 2 - ATM AAL5 PVC 3 - ATM Cell 4 - Ethernet VLAN 5 - Ethernet 6 - HDLC ( Cisco ) 7 - PPP Martini, et al. [Page 7] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 - VC ID length Length of the VC ID field in octets. If this value is 0, then it references all VCs using the specified group ID - Group ID An arbitrary 32 bit value which represents a group of VCs. - VC ID Identifies a particular VC. The interpretation of the identifier depends on the VC type: * Frame Relay A 32-bit value representing a 16-bit DLCI value 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 2 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Reserved | DLCI | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ * ATM AAL5 PVC A 32-bit value representing a 16-bit VPI, and a 16-bit VCI 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 2 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | VPI | VCI | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ * ATM Cell A 32-bit value representing a 16-bit VPI, and a 16-bit VCI 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 2 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | VPI | VCI | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Martini, et al. [Page 8] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 * Ethernet VLAN A 32 bit value representing 16bit vlan identifier 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 2 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Reserved | VLAN ID | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ * Ethernet A port identifier (details TBD). * HDLC ( Cisco ) A 16-bit port identifier (details TBD). * PPP A 16-bit port identifier (details TBD). 7. Security Considerations This document does not affect the underlying security issues of MPLS. 8. References [1] "LDP Specification", draft-ietf-mpls-ldp-06.txt, 10/5/99 [2] ITU-T Recommendation Q.933, and Q.922 Specification for Frame Mode Basic call control, ITU Geneva 1995 [3] "MPLS Label Stack Encoding", draft-ietf-mpls-label-encaps-07.txt, 9/13/99 [4] "IEEE 802.3ac-1998" IEEE standard specification. Martini, et al. [Page 9] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 9. Author Information Luca Martini Level 3 Communications, LLC. 1025 Eldorado Blvd. Broomfield, CO, 80021 e-mail: luca@level3.net Nasser El-Aawar Level 3 Communications, LLC. 1025 Eldorado Blvd. Broomfield, CO, 80021 e-mail: nna@level3.net Dimitri Stratton Vlachos Cisco Systems, Inc. 250 Apollo Drive Chelmsford, MA, 01824 e-mail: dvlachos@cisco.com Dan Tappan Cisco Systems, Inc. 250 Apollo Drive Chelmsford, MA, 01824 e-mail: tappan@cisco.com Eric Rosen Cisco Systems, Inc. 250 Apollo Drive Chelmsford, MA, 01824 e-mail: erosen@cisco.com Steve Vogelsang Laurel Networks, Inc. 2607 Nicholson Rd. Sewickley, PA 15143 e-mail: sjv@laurelnetworks.com Martini, et al. [Page 10] Internet Draft draft-martini-l2circuit-trans-mpls-01.txt May 2000 John Shirron Laurel Networks, Inc. 2607 Nicholson Rd. Sewickley, PA 15143 e-mail: sjv@laurelnetworks.com Martini, et al. [Page 11]