Using A Generic Associated Channel Label as a Virtual Circuit Connectivity Verification Channel Indicator

The Virtual Circuit Connectivity Verification (VCCV) protocol is specified in RFC 5085. This document specifies a new VCCV control channel (VCCV CC) type to be used with pseudowires (PW) carried over an MPLS network that do not use the PW Control Word (CW) . This new VCCV CC type uses the Generic Associated Channel Label (GAL) to distinguish VCCV packets from packets carrying user data. This new VCCV CC type provides compatibility with the method of MPLS Label Switched Path (LSP) Operations, Administration, and Maintenance (OAM) message identification, as used in MPLS Transport Profile (MPLS-TP) networks . VCCV currently specifies three CC types. VCCV CC Type 1 uses the PW Control Word (CW) to distinguish VCCV packets from packets carrying user data. VCCV CC Types 2 and 3 require IP encapsulation for OAM packets. This was not an issue when was designed, but is in conflict with the design goals of MPLS-TP which does not otherwise require the availability of IP. VCCV CC Type 2 is not applicable to Multi-Segment PWs (MS-PWs) . A MS-PW operating without the CW therefore has to use VCCV CC Type 3 which identifies VCCV packets on the basis of Time to Live (TTL) expiry. Whilst less of an issue with a single segment PW (SS-PW), on an MS-PW this requires accurately setting the TTL for expiry at the egress Terminating Provider Edge (T-PE) . In the event of an error in the setting of the PW Label Stack Entry (LSE) TTL, VCCV packets will not be received by the Terminating Provider Edge (T-PE) and may leak into the attachment circuit . The new VCCV CC type defined in this specification addresses these problems for PWs that do not use the CW. Note that mandating PWs use the PW CW is not a viable way to address this issue. This is because: PWs without the CW are already widely deployed. There is a significant deployment of existing hardware that does not support usage of the PW CW for some PW types. Some operators are concerned that the inclusion of the PW CW will increase the PW packet size.

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

When the PW CW is not used, the Type 4 MPLS VCCV Control Channel (CC) type defined in this section MAY be used. This is referred to as VCCV CC Type4 throughout the rest of this of this document. VCCV Type 4 uses the encapsulation shown in in which the presence of a GAL at the end of the MPLS label stack indicates that the packet carries a VCCV message.

The VCCV message body is preceded by a Generic Associated Channel Header as defined in , in which the Channel Type identifies the type and format of the OAM message carried in the VCCV message body. The GAL LSE MUST contain the GAL reserved label as defined in . The PW LSE is constructed according to the existing procedures that apply to the type of pseudowire that is in use. Note that the inclusion of a GAL following the PW LSE over a label switched path subject to Equal-Cost Multi-path (ECMP) load balancing can cause the OAM packet to take a different path through the network from the corresponding PW data packets. If that is not acceptable, then an alternative VCCV type needs to be used.

specifies that when the flow-aware transport (FAT) of pseudowires over an MPLS packet switched network has been signalled or configured, the Flow LSE MUST be present. It further specifies that "the flow label MUST NOT be an MPLS reserved label (values in the range 0..15) [RFC3032]", and that "If a flow LSE is present, it MUST be checked to determine whether it carries a reserved label. If it is a reserved label, the packet is processed according to the rules associated with that reserved label; otherwise, the LSE is discarded." This document specifies that if the flow-aware transport of pseudowires over an MPLS packet switched network has been signalled or configured then the presence of VCCV message is indicated by the use of a GAL in place of the flow LSE. This is consistent with , and the packet structure is identical to that shown in . Flow LSEs are inserted into a PW label stack in order to enable the distribution of the PW traffic among multiple equal cost MPLS paths. The use of GAL in place of the flow label will cause all OAM packets to take exactly one of these paths, and this path may be different from the paths taken by any of traffic flows. If this is not acceptable, then an alternative VCCV type needs be used.

When using VCCV CC Type 4 for MS-PWs, a PE transmitting the VCCV packet to a Switching PE (S-PE) MUST set the TTL to the appropriate value to expire at that S-PE. An S-PE that supports this specification MUST inspect PW packets that are received as a result of TTL expiry, and determine whether a GAL follows the PW LSE. If a GAL is present the S-PE then processes the VCCV packet. An S-PE that does not support this specification would be expected to reject as malformed a VCCV CC Type 4 packet that was received. This is because the S-PE would expect the PW LSE to be bottom of stack (the non FAT case) and for the LSE at bottom of stack not to be a reserved label (both the FAT and the non-FAT cases). An S-PE that did not make this reserved label check would then find that the first nibble following the label stack was 0x1 and not the expected start of an IP packet. It would hence be expected to also reject the packet. This update to the behaviour of S-PEs is therefore backwards compatible.

The VCCV capability advertisement MUST match the c-bit setting that is advertised in the PW FEC element . If the c-bit is set, indicating the use of the PW CW, then VCCV CC Type 4 MUST NOT be advertised. If the c-bit is not set, indicating that the PW CW is not in use, then equipment supporting this specification MUST advertise VCCV CC Type 4. Advertisement of VCCV CC Types 1 and 4 are therefore mutually exclusive. A PE supporting VCCV CC Type 4 MAY advertise other VCCV CC types as defined in . If the remote PE supports VCCV CC Type 4, and the PW CW is not in use, then for cases where multiple CC Types are advertised, the following precedence rules apply when choosing which CC Type to use: Type 4: GAL VCCV Control Channel. Type 2: MPLS Router Alert Label. Type 3: MPLS PW Label with TTL == 1. If the remote PE finds that VCCV CC Types 1 and 4 are both advertised, or that c-bit is set and VCCV CC Type 4 is advertised, then it should report the error to the operator through the management interface in use, and send a Label Release Message with a status code "VCCV Type Error".

Whilst the introduction of this additional VCCV CC type increases the number of VCCV CC types that the operator needs to manage, it addresses the issues with VCCV CC Types 2 and 3 described in . In the event of a misconfiguration of this VCCV CC type, the PW is taken out of service and the operator advised as described in . Attention is drawn to the possible absence of fate sharing between PW data packets and VCCV CC Type 4 packets described in and .

This document does not by itself raise any new security considerations beyond those described in and . provides detailed operational procedures which can be used to verify the MS-PW connectivity. In addition, the procedure specified in this document eliminates the possibility of packet leaking that can occur with VCCV Type 3.

IANA is requested to assign a new bit from the MPLS VCCV Control Channel (CC) Types registry in the PWE3-parameters name space in order to identify VCCV type 4. It is requested that Bit 3 be assigned to this purpose which would have a value of 0x08.

IANA is requested to assign a new Status Code from the Label Distribution Protocol (LDP) Parameters name space:

The authors wish to thank Alexander (Sasha) Vainshtein for his proposal to make the GAL and Flow labels mutually exclusive. This proposal led to a significant simplification of this design. The authors also thank Sasha, Matthew Bocci, Loa Andersson and Deborah Brungard for their review comments.