MPLS Generic Associated ChannelAlcatel-LucentVoyager Place, Shoppenhangers RoadMaidenheadBerksSL6 2PJUKmatthew.bocci@alcatel-lucent.comAlcatel-LucentRoute de VillejustNozay91620Francemartin.vigoureux@alcatel-lucent.comCiscoswallow@cisco.comCiscodward@cisco.comCiscostbryant@cisco.comJuniper Networksrahul@juniper.net
Routing
MPLS Working GroupMPLS-TPMPLSOAMGALG-AChThis document generalizes the applicability of the pseudowire (PW)
Associated Channel Header (ACH), enabling the realization of a control
channel associated to MPLS Label Switched Paths (LSPs) and MPLS Sections
in addition to MPLS pseudowires. In order to identify the presence of
this Associated Channel Header in the label stack, this document also
assigns one of the reserved MPLS label values to the Generic Associated
Channel Label (GAL), to be used as a label based exception
mechanism.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.There is a need for Operations, Administration and Maintenance (OAM)
mechanisms that can be used for fault detection, diagnostics,
maintenance and other functions on a pseudowire (PW) and a Label
Switched Path (LSP). These functions can be used between any two Label
Edge Routers (LERs) / Label Switching Router (LSRs) or Terminating
Provider Edge routers (T-PEs) / Switching Provider Edge routers (S-PEs)
along the path of an LSP or PW respectively . Some of these functions can
be supported using existing tools such as Virtual Circuit Connectivity
Verification (VCCV) , Bidirectional
Forwarding Detection for MPLS LSPs (BFD-MPLS) , LSP-Ping , or BFD-VCCV . However, a requirement has been
indicated to augment this set of maintenance functions, in particular
when MPLS networks are used for packet transport services and transport
network operations . Examples of these
functions include performance monitoring, automatic protection
switching, and support for management and signaling communication
channels. These tools MUST be applicable to, and function in essentially
the same manner (from an operational point of view) on MPLS PWs, MPLS
LSPs and MPLS Sections. They MUST also operate in-band on the PW or LSP
such that they do not depend on Packet Switched Network (PSN) routing or
on user traffic, and MUST also NOT depend on dynamic control plane
functions.VCCV can use an Associated Channel
Header (ACH) to provide a PW-associated control channel between a PW's
end points, over which OAM and other control messages can be exchanged.
This document generalizes the applicability of the ACH to enable the
same associated control channel mechanism to be used for Sections, LSPs
and PWs. The associated control channel thus generalized is known as the
Generic Associated Channel (G-ACh). The ACH, specified in RFC 4385 , may be used with additional code points to
support additional MPLS maintenance functions on the G-ACh.Generalizing the applicability of the ACH to LSPs and Sections also
requires a method to identify that a packet contains an ACH followed by
a non-service payload. Therefore, this document also defines a label
based exception mechanism that serves to inform an LSR (or LER) that a
packet it receives on an LSP or Section belongs to an associated control
channel. The label used for that purpose is one of the MPLS reserved
labels and is referred to as the GAL (G-ACh Label). The GAL mechanism is
defined to work together with the ACH for LSPs and MPLS Sections.RFC 4379 and BFD-MPLS define alert mechanisms that enable
an MPLS LSR to identify and process MPLS OAM packets when these are
encapsulated in an IP header. These alert mechanisms are based, for
example, on Time To Live (TTL) expiration and/or on the use of an IP
destination address in the range of 127/8 or
0:0:0:0:0:FFFF:127.0.0.0/104, respectively for IPv4 and IPv6. These
mechanisms are the default mechanisms for identifying MPLS OAM packets
when encapsulated in an IP header. However it may not always be possible
to use these mechanisms in some MPLS applications e.g., MPLS Transport
Profile (MPLS-TP) ,
particularly when IP based demultiplexing cannot be used. This document
defines a mechanism that is RECOMMENDED for identifying and
encapsulating MPLS OAM and other maintenance messages when IP based
mechanisms such as those used in and
are not available. Yet, this
mechanism MAY be used in addition to IP-based mechanisms.Note that, in this document, maintenance functions and packets should
be understood in the broad sense. That is, a set of maintenance and
management mechanisms that include OAM, Automatic Protection Switching
(APS), Signaling Communication Channel (SCC) and Management
Communication Channel (MCC) messages.Also note that the GAL and ACH are applicable to MPLS and PWs in
general. This document specifies general mechanism and uses MPLS-TP as
an example application. The application of the GAL and ACH to other
specific MPLS uses is outside the scope of this document.The editors gratefully acknowledge the contributions of Sami
Boutros, Italo Busi, Marc Lasserre, Lieven Levrau and Siva
SivabalanThis document defines a mechanism that provides a solution to the
extended maintenance needs of emerging applications for MPLS. It
creates a generic control channel mechanism that may be applied to
MPLS LSPs and Sections, while maintaining compatibility with the PW
associated channel. It also normalizes the use of the ACH for PWs in a
transport context, and defines a label based exception mechanism to
alert LERs/LSRs of the presence of an ACH after the bottom of the
stack.This document defines the encapsulation header for Sections, LSPs,
and PWs associated control channel messages.It does not define how associated control channel capabilities are
signaled or negotiated between LERs/LSRs or PEs, or the operation of
various OAM functions.This document does not deprecate existing MPLS and PW OAM
mechanisms.ACH: Associated Channel HeaderG-ACh: Generic Associated ChannelGAL: G-ACh LabelG-ACh packet: Any packet containing a message belonging to a
protocol that is carried on a PW, LSP or MPLS Section associated
control channel. Examples include maintenance protocols such as OAM
functions, signaling communications or management communications.The terms 'Section' and 'Concatenated Segment' are defined in as follows (note that
the terms 'Section' and ''Section Layer Network' are synonymous):Concatenated Segment: A serial-compound link connection as defined
in . A concatenated segment is a contiguous
part of an LSP or multi-segment PW that comprises a set of segments
and their interconnecting nodes in sequence.Section Layer Network: A section is a server layer (which may be
MPLS-TP or a different technology) which provides for encapsulation
and OAM of a client layer network. A section layer may provide for
aggregation of multiple MPLS-TP clients. Note that G.805 defines the section layer as one of the two
layer networks in a transmission media layer network. The other layer
network is the physical media layer network.VCCV defines three Control Channel
(CC) Types that may be used to exchange OAM messages through a PW: CC
Type 1 uses an ACH and is referred to as "In-band VCCV"; CC Type 2 uses
the MPLS Router Alert Label to indicate VCCV packets and is referred to
as "Out of Band VCCV"; CC Type 3 uses the TTL to force the packet to be
processed by the targeted router control plane and is referred to as
"MPLS PW Label with TTL == 1".The use of the ACH, previously limited to PWs, is here generalized
to also apply to LSPs and to Sections. Note that for PWs, the PWE3
control word MUST be present in the
encapsulation of user packets when the ACH is used to realize the
associated control channel.The ACH used by CC Type 1 is depicted in figure below:In the above figure, the first nibble is set to 0001b to indicate a
control channel associated with a PW, an LSP or a Section. The Version
field is set to 0, as specified in RFC 4385 . Bits 8 to 15 of the ACH are reserved and
MUST be set to 0 and ignored on reception. Bits 16 to 31 are used to
encode the possible Channel Types.Note that VCCV also includes
mechanisms for negotiating the Control Channel and Connectivity
Verification (i.e., OAM function) Types between PEs. It is anticipated
that similar mechanisms will be applied to LSPs. Such application will
require further specification. However, such specification is beyond
the scope of this document.The G-ACh MUST NOT be used to transport user traffic.The Channel Type field indicates the type of message carried on the
associated control channel e.g., IPv4 or IPv6 if IP demultiplexing is
used for messages sent on the associated control channel, or OAM or
other maintenance function if IP demultiplexing is not used. For
associated control channel packets where IP is not used as the
multiplexer, the Channel Type indicates the specific protocol carried
in the associated control channel.Values for the Channel Type field currently used for VCCV are
specified elsewhere e.g., in RFC 4446
and RFC 4385 . Additional Channel Type
values and the associated maintenance functionality will be defined in
other documents. Each document, specifying a protocol solution relying
on the ACH, MUST also specify the applicable Channel Type field
value.Note that these values are allocated from the PW Associated Channel
Type registry , but this document
modifies the existing policy to accommodate a level of
experimentation. See for further
details.In some applications of the generalized associated control channel it
is necessary to include one or more ACH TLVs to provide additional
context information to the G-ACh packet. One use of these ACH TLVs might
be to identify the source and/or intended destination of the associated
channel message. However, the use of this construct is not limited to
providing addressing information nor is the applicability restricted to
transport network applications.If the G-ACh message MAY be preceded by one or more ACH TLVs, then
this MUST be explicitly specified in the definition of an ACH Channel
Type. If the ACH Channel Type definition does state that one or more ACH
TLVs MAY precede the G-ACh message, an ACH TLV Header MUST follow the
ACH. If no ACH TLVs are required in a specific associated channel
packet, but the Channel Type nevertheless defines that ACH TLVs MAY be
used, an ACH TLV Header MUST be present but with a length field set to
zero to indicate that no ACH TLV follow this header.If an ACH Channel Type specification does not explicitly specify that
ACH TLVs MAY be used, then the ACH TLV Header MUST NOT be used.This section defines and describes the structure of an ACH payload
when an ACH TLV Header is present. The structure of ACH TLVs that MAY
follow an ACH TLV Header is defined and described in the following
sections.The following figure () shows the
structure of a G-ACh packet payload.The ACH TLV Header defines the length of the set of ACH TLVs that
follow.The Length field specifies the length in octets of the complete set
of TLVs including sub-TLVs that follow the ACH TLV header. A length of
zero indicates that no ACH TLV follow this header. Note that no
padding is required for the set of ACH TLVs.The Reserved field is for future use and MUST be set to zero on
transmission and ignored on reception.An ACH TLV consists of a 16-bit Type field, followed by a 16-bit
Length field which specifies the number of octets of the Value field
which follows the Length field. This 32-bit word is followed by zero
or more octets of Value information. The format and semantics of the
Value information are defined by the TLV Type as recorded in the TLV
Type registry. See for further details.
Note that the Value field of ACH TLVs MAY contain sub-TLVs. Note that
no padding is required for individual TLVs or sub-TLVs.Generalizing the associated control channel mechanism to LSPs and
Sections also requires a method to identify that a packet contains an
ACH followed by a non-service payload. This document specifies that a
label is used for that purpose and calls this special label the G-ACh
Label (GAL). One of the reserved label values defined in RFC 3032 is assigned for this purpose. The value of the
label is to be allocated by IANA; this document suggests the value
13.The GAL provides an alert based exception mechanism to:differentiate specific packets (i.e., G-ACh packets) from others,
such as user-plane ones,indicate that the ACH appears immediately after the bottom of the
label stack.The GAL MUST only be used where both these purposes apply.RFC 4379 and BFD-MPLS define alert mechanisms that enable
an MPLS LSR to identify and process MPLS OAM packets when these are
encapsulated in an IP header. These alert mechanisms are based, for
example, on Time To Live (TTL) expiration and/or on the use of an IP
destination address in the range of 127/8 or
0:0:0:0:0:FFFF:127.0.0.0/104, respectively for IPv4 and IPv6.These mechanisms are the default mechanisms for identifying MPLS
OAM packets when encapsulated in an IP header although the mechanism
defined in this document MAY also be used.In MPLS-TP, the GAL MUST be used with packets on a G-ACh on LSPs,
Concatenated Segments of LSPs, and with Sections, and MUST NOT be used
with PWs. It MUST always be at the bottom of the label stack (i.e., S
bit set to 1). However, in other MPLS environments, this document
places no restrictions on where the GAL may appear within the label
stack or its use with PWs. Where the GAL is at the bottom of the label
stack (i.e. S bit set to 1) then it MUST always be followed by an
ACH.The GAL MUST NOT appear in the label stack when transporting normal
user-plane packets. Furthermore, when present, the GAL MUST NOT appear
more than once in the label stack.A receiving LSR, LER or PE MUST NOT forward a G-ACh packet to
another node based on the GAL label.The Traffic Class (TC) field (formerly known as the EXP field) of
the Label Stack Entry (LSE) containing the GAL follows the
definition and processing rules specified and referenced in .The Time-To-Live (TTL) field of the LSE that contains the GAL
follows the definition and processing rules specified in .The following figure () depicts
two LERs (A and D) and two LSRs (B and C) for a given LSP which is
established from A to D and switched in B and C.In this example, a G-ACh exists on the LSP that extends between
LERs A and D, via LSRs B and C. Only A and D may initiate the
generation of G-ACh packets. A, B, C and D may also originate and
process G-ACh packets.The following figure ()
depicts the format of a MPLS-TP G-ACh packet when used for an
LSP.Note that it is possible that the LSP may be tunneled in
another LSP (e.g., if a MPLS Tunnel exists between B and C), and
as such other LSEs may be present in the label stack.To send a maintenance message on the LSP associated control
channel, the LER (A) generates a G-ACh message, to which it MAY
prepend an ACH TLV Header and appropriate ACH TLVs, adds an ACH to
which it, pushes a GAL LSE and finally the LSP Label LSE.The TTL field of the GAL LSE MUST be set to at least 1. The
exact value of the TTL is application specific. See Section
4.2.1 for definition and processing rules.The S bit of the GAL MUST be set according to its position
in the label stack (see Section 4.2).The setting of the TC field of the GAL is application
specific. See Section 4.2.1 for definition and processing
rules.The G-ACh message, the ACH or the GAL SHOULD NOT be
modified towards the targeted destination. Upon reception of the
labeled packet, the targeted destination, after having checked
both the LSP Label and GAL LSEs fields, SHOULD pass the whole
packet to the appropriate processing entity.The following figure ()
depicts an example of an MPLS Section.With regard to the MPLS Section, a G-ACh exists between A and
Z. Only A and Z can insert, extract or process packets on this
G-ACh.The following figure () depicts the format of a
G-ACh packet when used for an MPLS Section. The GAL MAY provide
the exception mechanism for a control channel in its own right
without being associated with a specific LSP, thus providing
maintenance related communications across a specific link
interconnecting two LSRs. In this case, the GAL is the only label
in the stack.To send a G-ACh message on a control channel associated to the
Section, the head-end LSR (A) of the Section generates a G-ACh
message, to which it MAY prepend an ACH TLV Header and appropriate
ACH TLVs, adds an ACH to which it pushes a GAL LSE.The TTL field of the GAL MUST be set to at least 1. The
exact value of the TTL is application specific. See Section
4.2.1 for definition and processing rules.The S bit of the GAL MUST be set according to its position
in the label stack. (see Section 4.2).The setting of the TC field of the GAL is application
specific. See Section 4.2.1 for definition and processing
rules.The G-ACh message, the ACH and the GAL SHOULD NOT be modified
towards the tail-end LSR (Z). Upon reception of the G-ACh packet,
the tail-end LSR (Z), after having checked the GAL LSE fields,
SHOULD pass the whole packet to the appropriate processing
entity.RFC 3429 describes the assignment of
one of the reserved label values, defined in RFC 3032 , to the 'OAM Alert Label' that is used by
user-plane MPLS OAM functions for the identification of MPLS OAM
packets. The value of 14 is used for that purpose.Both this document and RFC 3429
therefore describe the assignment of reserved label values for similar
purposes. The rationale for the assignment of a new reserved label can
be summarized as follows:Unlike the mechanisms described and referenced in RFC 3429
, G-ACh messages will not reside
immediately after the GAL but instead behind the ACH, which itself
resides after the bottom of the label stack.The set of maintenance functions potentially operated in the
context of the G-ACh is wider than the set of OAM functions
referenced in RFC 3429 .It has been reported that there are existing implementations
and running deployments using the 'OAM Alert Label' as described
in RFC 3429 . It is therefore not
possible to modify the 'OAM Alert Label' allocation, purpose or
usage. Nevertheless, it is RECOMMENDED that no further OAM
extensions based on 'OAM Alert Label' (Label 14) usage be
specified or developed.Procedures for handling a packet received with an invalid incoming
label are specified in RFC 3031.An LER, LSR or PE MUST discard received associated channel packets on
which all of the MPLS or PW labels have been popped if any one of the
following conditions is true:It is not capable of processing packets on the Channel Type
indicated by the ACH of the received packet.It has not, through means outside the scope of this document,
indicated to the sending LSR, LER or PE that it will process
associated channel packets on the Channel Type indicated by the ACH
of the received packet.The packet is received on an Experimental Channel Type that is
locally disabled.If the ACH was indicated by the presence of a GAL, and the first
nibble of the ACH of the received packet is not 0b0001.The ACH version is not recognized.In addition, it MAY increment an error counter and MAY also
issue a system and/or SNMP notification.The congestion considerations detailed in RFC 5085 apply.The security considerations for the associated control channel are
described in RFC 4385 . Further security
considerations MUST be described in the relevant associated channel type
specification.RFC 5085 provides data plane related
security considerations. These also apply to a G-ACh, whether the alert
mechanism uses a GAL or only an ACH.This document requests that IANA allocates a label value, to the GAL,
from the pool of reserved labels in the "Multiprotocol Label Switching
Architecture (MPLS) Label Values" registry, and suggests this value to
be 13.Channel Types for the Associated Channel Header are allocated from
the IANA "PW Associated Channel Type" registry . The PW Associated Channel Type registry is
currently allocated based on the IETF consensus process, described in
. This allocation process was chosen based
on the consensus reached in the PWE3 working group that pseudowire
associated channel mechanisms should be reviewed by the IETF and only
those that are consistent with the PWE3 architecture and requirements
should be allocated a code point.However, a requirement has emerged (see ) to allow for
optimizations or extensions to OAM and other control protocols running
in an associated channel to be experimented without resorting to the
IETF standards process, by supporting experimental code points. This
would prevent code points used for such functions from being used from
the range allocated through the IETF standards and thus protects an
installed base of equipment from potential inadvertent overloading of
code points. In order to support this requirement, this document
requests that the code point allocation scheme for the PW Associated
Channel Type be changed as follows:0 - 32751 : IETF Consensus32760 - 32767 : ExperimentalCode points in the experimental range MUST be used according to the
guidelines of RFC 3692 . Functions using
experimental G-ACh code points MUST be disabled by default. The Channel
Type value used for a given experimental OAM function MUST be
configurable, and care MUST be taken to ensure that different OAM
functions that are not inter-operable are configured to use different
Channel Type values.The PW Associated Channel Type registry needs to be updated to
include a column indicating whether the ACH is followed by a ACH TLV
header (Yes/No). There are two ACH Channel Type code-points currently
assigned and in both cases no ACH TLV header is used. Thus the new
format of the PW Channel Type registry is:IANA is requested create a new registry
called the Associated Channel Header TLV Registry. The allocation policy
for this registry is IETF consensus. This registry MUST record the
following information. There are no initial entries.The authors would like to thank Malcolm Betts, ITU-T Study Group 15,
and all members of the teams (the Joint Working Team, the MPLS
Interoperability Design Team in IETF and the MPLS-TP Ad-Hoc Team in
ITU-T) involved in the definition and specification of the MPLS
Transport Profile.Generic Functional Architecture of Transport Networks