LMP SONET/SDH Test: Synchronous Optical Network (SONET)/Synchronous Digital Hierarchy (SDH) Encoding for Link Management Protocol (LMP) Test Messages
How LMP link verification works on SONET and SDH links, using trace bytes and new trace messages.
Lang, J. and D. Papadimitriou, "Synchronous Optical Network (SONET)/Synchronous Digital Hierarchy (SDH) Encoding for Link Management Protocol (LMP) Test Messages", RFC 4207, DOI 10.17487/RFC4207, October 2005, <https://www.rfc-editor.org/info/rfc4207>.
@misc{rfc4207,
series = {Request for Comments},
number = 4207,
howpublished = {RFC 4207},
publisher = {RFC Editor},
doi = {10.17487/RFC4207},
url = {https://www.rfc-editor.org/info/rfc4207},
author = {J. Lang and D. Papadimitriou},
title = {{Synchronous Optical Network (SONET)/Synchronous Digital Hierarchy (SDH) Encoding for Link Management Protocol (LMP) Test Messages}},
pagetotal = 15,
year = 2005,
month = oct,
} RFC text
Contents
Network Working Group J. Lang
Request for Comments: 4207 Sonos, Inc.
Category: Standards Track D. Papadimitriou
Alcatel
October 2005
Synchronous Optical Network (SONET)/Synchronous Digital Hierarchy (SDH)
Encoding for Link Management Protocol (LMP) Test Messages
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 (2005).
Abstract
This document details the Synchronous Optical Network
(SONET)/Synchronous Digital Hierarchy (SDH) technology-specific
information needed when sending Link Management Protocol (LMP) test
messages.
1. Introduction
For scalability purposes, multiple physical resources that
interconnect Label Switching Routers (LSRs) can be combined to form a
single traffic engineering (TE) link for the purposes of path
computation and signaling. These resources may represent one or more
physical links that connect the LSRs, or they may represent a Label
Switched Path (LSP) if LSP hierarchy [RFC4206] is used. The
management of TE links is not restricted to in-band messaging, but
instead can be done using out-of-band techniques.
The Link Management Protocol (LMP) [RFC4204] has been developed as
part of the Generalized MPLS (GMPLS) protocol suite to manage TE
links. LMP currently consists of four main procedures, of which the
first two are mandatory and the last two are optional:
1. Control channel management
2. Link property correlation
3. Link verification
4. Fault management
Control channel management is used to establish and maintain control
channel connectivity between adjacent nodes. This is done using a
Config message exchange followed by a lightweight keep-alive message
exchange. Link property correlation is used to aggregate multiple
data links into a single TE Link and to synchronize the link
properties. Link verification is used to verify the physical
connectivity of the data links and to exchange the Interface_Ids of
the data links. Fault management is primarily used to suppress
alarms and to localize failures in both opaque and transparent
networks. When LMP is used with SONET/SDH, however, the fault
management procedures may not be needed as existing SONET/SDH
mechanisms can be used.
In this document, the SONET/SDH technology-specific information for
LMP is defined. Specifically, the SONET/SDH test procedures used for
link verification and link property correlation are detailed. These
procedures include the trace correlation transport mechanism (defined
for J0, J1, J2) that supports a separation of the transport and
control plane identifiers. The latter procedure requires a new trace
monitoring function that is discussed in this document. Once the
data links have been verified, they can be grouped to form TE links.
2. Terminology
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 [RFC2119].
The reader is assumed to be familiar with the terminology in
[RFC4204], [G.707], and [T1.105]. The following abbreviations are
used in this document:
CRC-N: Cyclic Redundancy Check-N.
DCC: Data communications channel.
LOVC: Lower-order virtual container.
HOVC: Higher-order virtual container.
MS: Multiplex section.
MSOH: Multiplex section overhead.
POH: Path overhead.
RS: Regenerator section.
RSOH: Regenerator section overhead.
SDH: Synchronous digital hierarchy.
SOH: Section overhead.
SONET: Synchronous Optical Network.
STM(-N): Synchronous Transport Module (-N) (SDH).
STS(-N): Synchronous Transport Signal-Level N (SONET).
VC-n: Virtual Container-n (SDH).
VTn: Virtual Tributary-n (SONET).
3. Verifying Link Connectivity
In [RFC4204], a link verification procedure is defined whereby Test
messages are transmitted in-band over the data links. This is used
for data plane discovery, Interface_Id exchange (Interface_Ids are
used in GMPLS signaling, either as port labels [RFC3471] or component
link identifiers [RFC4201], depending on the configuration), and
physical connectivity verification. Multiple data links can be
verified using a single verification procedure; the correlation is
done using the Verify_Id that is assigned to the procedure.
As part of the link verification procedure, a BeginVerify message
exchange is used to agree upon parameters for the Test procedure.
This can be initiated by sending a BeginVerify message over the
control channel. This message includes a BEGIN_VERIFY object that
contains a number of fields specifying, among other things, the
transmission (bit) rate, encoding type, and transport mechanisms for
the Test Messages. If the remote node receives a BeginVerify message
and is ready to begin the procedure, it sends a BeginVerifyAck
message specifying the desired transport mechanism for the Test
messages. The remote node also assigns a Verify_Id to the procedure
and includes it in the BeginVerifyAck message.
The transmission rate of the data link over which the Test Messages
will be transmitted is represented in IEEE floating-point format
using a 32-bit number field and expressed in bytes per second. See
[RFC3471] for values defined for SONET/SDH.
The encoding type identifies the encoding supported by an interface.
The defined encoding is consistent with the LSP Encoding Type as
defined in [RFC3471]. For SONET/SDH, this value must equal the value
given for "SDH ITU-T G.707/SONET ANSI T1.105".
The transport mechanism is defined using the Verify Transport
Mechanism bit mask. The scope of this bit mask is restricted to the
link encoding type. Multiple bits may be set when this field is
included in the BeginVerify message; however, only one bit may be set
when it is included in the BeginVerifyAck message.
In the following subsection, the various options for Verify Transport
Mechanism are defined when the encoding is SONET/SDH. The trace
correlation transport mechanism (defined for J0, J1, J2) supports a
separation of the transport and control plane identifiers.
3.1. Verify Transport Mechanism
This field is 16 bits in length.
In this document, the flags for SONET/SDH encoding are defined. Note
that all values are defined in network byte order (i.e., big-endian
byte order).
0x0001: Reserved
0x0002 DCCS: Test Message over the Section/RS DCC
Capable of transmitting Test Messages using the DCC
Section/RS Overhead bytes with bit-oriented High-Level
Data Link Control (HDLC) framing format [RFC1662].
The Test Message is sent as defined in [RFC4204].
0x0004 DCCL: Test Message over the Line/MS DCC
Capable of transmitting Test Messages using the DCC
Line/MS Overhead bytes with bit-oriented HDLC framing
format [RFC1662].
The Test Message is sent as defined in [RFC4204].
0x0008 J0-trace: J0 Section Trace Correlation
Capable of transmitting SONET/SDH Section/RS trace over
J0 Section/RS overhead byte as defined in [T1.105] and
[G.707].
The Test Message is not transmitted using the J0 bytes
(i.e., over the data link), but is sent over the control
channel and correlated for consistency to the received
J0 pattern.
In order to get the mapping between the Interface_Id
over which the J0 Test Message is sent and the J0
pattern sent in-band, the transmitting node must provide
the correlation between this pattern and the J0 Test
Message. This correlation is done using the TRACE
object as defined in Section 4.
The format of the Test Message is as follows:
<Test Message> ::=<Common Header> <LOCAL_INTERFACE_ID>
<VERIFY_ID> <TRACE>
0x0010: Reserved
0x0020: Reserved
0x0040 J1-trace: J1 Path Trace Correlation
Capable of transmitting SONET/SDH STS SPE/HOVC Path
trace over J1 Path overhead byte as defined in [T1.105]
and [G.707].
The Test Message is not transmitted using the J1 bytes
(i.e., over the data link), but is sent over the control
channel and correlated for consistency to the received
J1 pattern.
In order to get the mapping between the Interface_Id
over which the J1 Test Message is sent and the J1
pattern sent in-band, the transmitting node must provide
the correlation between this pattern and the J1 Test
Message. This correlation is done using the TRACE
object as defined in Section 4.
The Test Message format is identical to that defined
above in J0-trace.
0x0080 J2-trace: J2 Path Trace Correlation
Capable of transmitting SONET/SDH VT SPE/LOVC Path trace
over J2 Path overhead byte as defined in [T1.105] and
[G.707].
The Test Message is not transmitted using the J2 bytes
(i.e., over the data link), but is sent over the control
channel and correlated for consistency to the received
J2 pattern.
In order to get the mapping between the Interface_Id
over which the J2 Test Message is sent and the J2
pattern sent in-band, the transmitting node must provide
the correlation between this pattern and the J2 Test
Message. This correlation is done using the TRACE
object as defined in Section 4.
The Test Message format is identical to that defined
above in J0-trace.
4. Trace Monitoring
The trace monitoring features described in this section allow a node
to do trace monitoring by using the SONET/SDH capabilities.
o A node may request its neighbor (the remote node) to monitor a
link for a specific pattern in the overhead using the
TraceMonitor Message. An example of this overhead is the SONET
Section Trace message transmitted in the J0 byte. If the
actual trace message does not match the expected trace message,
the remote node MUST report the mismatch condition.
o A node may request the value of the current trace message on a
given data link using the TraceReq Message.
o A node may request a remote node to send a specific trace
message over a data link using the InsertTrace Message.
4.1.1. TraceMonitor Message
The TraceMonitor message (Message Type 21) is sent over the
control channel and is used to request the remote node to monitor a
data link for a specific trace value. This value is inserted in the
<TRACE> object. The format of the TraceMonitor message is as
follows:
<TraceMonitor Message> ::= <Common Header> <MESSAGE_ID>
<LOCAL_INTERFACE_ID> <TRACE>
The above transmission order SHOULD be followed.
The remote node MUST respond to a TraceMonitor message with either a
TraceMonitorAck or TraceMonitorNack Message.
4.1.1.1. TRACE Object Class
Class = 21
o C-Type = 1, Trace
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|N| C-Type | Class | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Trace Type | Trace Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
// Trace Message //
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Trace Type: 16 bits
The type of the trace message. The following values are defined.
All other values are reserved.
1 = SONET Section Trace (J0 Byte)
2 = SONET Path Trace (J1 Byte)
3 = SONET Path Trace (J2 Byte)
4 = SDH Section Trace (J0 Byte)
5 = SDH Path Trace (J1 Byte)
6 = SDH Path Trace (J2 Byte)
Trace Length: 16 bits
This is the length in bytes of the trace message (as specified by
the Trace Type).
Trace Message:
This is the value of the expected message to be received in-band.
The valid length and value combinations are determined by the
specific technology: for SONET see [T1.105] and for SDH see
[G.707]. The message MUST be padded with zeros to a 32-bit
boundary, if necessary. Trace Length does not include padding
zeroes.
This object is nonnegotiable.
4.1.2. TraceMonitorAck Message
The TraceMonitorAck message (Message Type 22) is used to acknowledge
receipt of the TraceMonitor message and indicate that all of the
TRACE Objects in the TraceMonitor message have been received and
processed correctly (i.e., no Trace Mismatch).
The format is as follows:
<TraceMonitorAck Message> ::= <Common Header> <MESSAGE_ID_ACK>
The above transmission order SHOULD be followed.
The MESSAGE_ID_ACK object is defined in [RFC4204]. The contents of
the MESSAGE_ID_ACK object MUST be obtained from the TraceMonitor
message being acknowledged.
4.1.3. TraceMonitorNack Message
The TraceMonitorNack message (Message Type 23) is used to acknowledge
receipt of the TraceMonitor message and indicate that the TRACE
Object in the TraceMonitor message was not processed correctly. This
could be because the trace monitoring requested is not supported or
there was an error in the TRACE object value(s).
The format is as follows:
<TraceMonitorNack Message> ::= <Common Header> <MESSAGE_ID_ACK>
<ERROR_CODE>
The above transmission order SHOULD be followed.
The MESSAGE_ID_ACK and ERROR_CODE objects are defined in [RFC4204].
The contents of the MESSAGE_ID_ACK object MUST be obtained from the
TraceMonitor message being acknowledged.
If the Trace type is not supported, the ERROR_CODE MUST indicate
"Unsupported Trace Type" defined in Section 4.1.3.1.
If the TRACE object was not equal to the value seen in the trace, the
TraceMonitorNack message MUST include the ERROR_CODE indicating
"Invalid Trace Message". The TraceMismatch message (see Section
4.1.4) SHOULD NOT be sent as a result of the mismatch.
The TraceMonitorNack message uses a new ERROR_CODE C-Type defined in
Section 4.1.3.1.
4.1.3.1. ERROR_CODE Class
C-Type = 3, TRACE_ERROR
The following new error code bit-values are defined:
0x01 = Unsupported Trace Type
0x02 = Invalid Trace Message
All other values are Reserved.
Multiple bits may be set to indicate multiple errors.
This Object is nonnegotiable.
4.1.4. TraceMismatch Message
The TraceMismatch message (Message Type 24) is sent over the control
channel and is used to report a trace mismatch on a data link for
which trace monitoring was requested. The format is as follows:
<TraceMismatch message> ::= <Common Header> <MESSAGE_ID>
<LOCAL_INTERFACE_ID>
[<LOCAL_INTERFACE_ID> ...]
The above transmission order SHOULD be followed.
A neighboring node that receives a TraceMismatch message MUST respond
with a TraceMismatchAck message.
The LOCAL_INTERFACE_ID object is defined in [RFC4204]. The
LOCAL_INTERFACE_ID in this message is the local Interface Id of the
data link that has a trace mismatch. A trace mismatch for multiple
LOCAL_INTERFACE_IDs may be reported in the same message.
4.1.5. TraceMismatchAck Message
The TraceMismatchAck message (Message Type 25) is used to acknowledge
receipt of a TraceMismatch message. The format is as follows:
<TraceMismatchAck Message> ::= <Common Header> <MESSAGE_ID_ACK>
The MESSAGE_ID_ACK object is defined in [RFC4204]. The contents of
the MESSAGE_ID_ACK object MUST be obtained from the TraceMismatch
message being acknowledged.
4.1.6. TraceReq Message
The TraceReq message (Message Type 26) is sent over the control
channel and is used to request the current trace value of a data
link.
<TraceReq Message> ::= <Common Header> <MESSAGE_ID>
<LOCAL_INTERFACE_ID> <TRACE_REQ>
The above transmission order SHOULD be followed.
The format of the TRACE_REQ object is as follows:
Class = 22
O C-Type = 1, TraceReq
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|N| C-Type | Class | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Trace Type | (Reserved) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Trace Type: 16 bits
Defined in Section 4.1.1.1.
Reserved: 16 bits
This field MUST be set to zero when sent and ignored when
received
4.1.7. TraceReport Message
The TraceReport message (Message Type 27) is sent over the control
channel after receiving a TraceReq message.
<TraceReport Message> ::= <Common Header> <MESSAGE_ID_ACK> <TRACE>
The TraceReport message MUST include a TRACE Object (as described in
Section 4.1.1.1) for the requested data link.
The MESSAGE_ID_ACK object is defined in [RFC4204]. The contents of
the MESSAGE_ID_ACK object MUST be obtained from the TraceReq message
being acknowledged.
4.1.8. TraceReqNack Message
The TraceReqNack message (Message Type 28) is sent over the control
channel after receiving a TraceReq message.
<TraceReqNack Message> ::= <Common Header> <MESSAGE_ID_ACK>
<ERROR_CODE>
The above transmission order SHOULD be followed.
The MESSAGE_ID_ACK object is defined in [RFC4204]. The contents of
the MESSAGE_ID_ACK object MUST be obtained from the TraceReq message
being acknowledged.
The TraceReqNack message MUST include an ERROR_CODE Object (as
defined in Section 4.1.3.1) for the requested data link.
4.1.9. InsertTrace Message
The InsertTrace message (Message Type 29) is sent over the control
channel and is used to request a remote node to send a specific trace
message over a data link (this assumes that the remote knows the
mapping between the local and remote interface_Ids before fulfilling
such request).
The format is as follows:
<InsertTrace Message> ::= <Common Header> <MESSAGE_ID>
<LOCAL_INTERFACE_ID> <TRACE>
The above transmission order SHOULD be followed.
A node that receives an InsertTrace message MUST respond with either
an InsertTraceAck or an InsertTraceNack Message.
Once the InsertTraceAck message is received, the TraceMismatch
message (see Section 4.1.4) is used to indicate a trace mismatch has
occurred.
The MESSAGE_ID_object is defined in [RFC4204].
4.1.10. InsertTraceAck Message
The InsertTraceAck message (Message Type 30) is used to acknowledge
receipt of the InsertTrace message and indicate that the TRACE Object
in the InsertTrace message has been received and processed correctly
(i.e., no Trace Mismatch). The format is as follows:
<InsertTraceAck Message> ::= <Common Header> <MESSAGE_ID_ACK>
The MESSAGE_ID_ACK object is defined in [RFC4204]. The contents of
the MESSAGE_ID_ACK object MUST be obtained from the InsertTrace
message being acknowledged.
4.1.11. InsertTraceNack Message
The InsertTraceNack message (Message Type 31) is used to acknowledge
receipt of the InsertTrace message and to indicate that the TRACE
Object in the InsertTrace message was not processed correctly. This
could be because the trace monitoring requested is not supported or
there was an error in the value.
The format is as follows:
<InsertTraceNack Message> ::= <Common Header> <MESSAGE_ID_ACK>
<ERROR_CODE>
The above transmission order SHOULD be followed.
The MESSAGE_ID_ACK object is defined in [RFC4204].
The InsertTraceNack message MUST include an ERROR_CODE Object (as
defined in Section 4.1.3.1) for the requested data link.
5. Security Considerations
LMP message security uses IPsec as described in [RFC4204]. This
document introduces no other new security considerations not covered
in [RFC4204].
6. IANA Considerations
LMP [RFC4204] defines the following name spaces and how IANA can make
assignments in those namespaces:
- LMP Message Type.
- LMP Object Class.
- LMP Object Class type (C-Type) unique within the Object Class.
- LMP Sub-object Class type (Type) unique within the Object Class.
This memo introduces the following new assignments:
LMP Message Type:
o TraceMonitor message (Message type = 21)
o TraceMonitorAck message (Message type = 22)
o TraceMonitorNack message (Message type = 23)
o TraceMismatch message (Message type = 24)
o TraceMismatchAck message (Message type = 25)
o TraceReq message (Message type = 26)
o TraceReport message (Message type = 27)
o TraceReqNack message (Message type = 28)
o InsertTrace message (Message type = 29)
o InsertTraceAck message (Message type = 30)
o InsertTraceNack message (Message type = 31)
LMP Object Class name space and Class type (C-Type):
o TRACE Class name (21)
- Type 1 (C-Type = 1)
o TRACE REQ Class name (22)
- Type 1 (C-Type = 1)
7. References
7.1. Normative References
[RFC4201] Kompella, K., Rekhter, Y., and L. Berger, "Link Bundling
in MPLS Traffic Engineering (TE)", RFC 4201, October
2005.
[G.707] ITU-T Recommendation G.707, "Network node interface for
the synchronous digital hierarchy (SDH)," October 2000.
[RFC4204] Lang, J., Ed., "Link Management Protocol (LMP)", RFC
4204, October 2005.
[RFC1662] Simpson, W., "PPP in HDLC-like Framing", STD 51, RFC
1662, July 1994.
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC3471] Berger, L., "Generalized Multi-Protocol Label Switching
(GMPLS) Signaling Functional Description", RFC 3471,
January 2003.
[T1.105] T1.105, "Revised Draft T105 SONET Base Standard," January
2001.
7.2. Informative References
[RFC4206] Kompella, K., and Y. Rekhter, "Label Switched Paths (LSP)
Hierarchy with Generalized Multi-Protocol Label Switching
(GMPLS) Traffic Engineering (TE)", RFC 4206, October
2005.
8. Acknowledgements
The authors would like to thank Bernard Sales, Emmanuel Desmet, Gert
Grammel, Jim Jones, Stefan Ansorge, John Drake, and James Scott for
their many contributions to this document.
We would also like to thank Greg Bernstein and Michiel van Everdingen
for their insightful comments and for acting with a strong
combination of toughness, professionalism, and courtesy.
Authors' Addresses
Jonathan P. Lang
Sonos, Inc.
223 E. De La Guerra St.
Santa Barbara, CA 93101
EMail: jplang@ieee.org
Dimitri Papadimitriou
Alcatel
Francis Wellesplein 1
B-2018 Antwerpen, Belgium
EMail: dimitri.papadimitriou@alcatel.be
Full Copyright Statement
Copyright (C) The Internet Society (2005).
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 currently provided by the
Internet Society.
AI ELI5
On this page
§1What This Memo Adds to LMP
RFC 4207 is a companion to RFC 4204, the Link Management Protocol (LMP). LMP has four procedures: control channel management, link property correlation, link verification and fault management. The first two are mandatory and the last two are optional. This memo supplies the SONET/SDH technology-specific information for the link verification and link property correlation procedures.
The RFC also notes that when LMP is used with SONET/SDH, the fault management procedures may not be needed, because existing SONET/SDH mechanisms can be used.
The central idea is the trace correlation transport mechanism, defined for the J0, J1 and J2 overhead bytes. It separates the transport plane identifiers from the control plane identifiers, and it needs a new trace monitoring function that the memo defines.
LMP already has a way to test cables. This memo adds the special instructions for one kind of cable, SONET/SDH fiber.
§3Verify Transport Mechanisms
In RFC 4204, a BeginVerify message agrees on test parameters, among them the bit rate, the encoding type and the transport mechanism. The data rate is a 32-bit IEEE floating-point number in bytes per second. For SONET/SDH the encoding type must equal the value for "SDH ITU-T G.707/SONET ANSI T1.105" from RFC 3471.
The Verify Transport Mechanism is a 16-bit bit mask. Several bits can be set in BeginVerify, but only one may be set in BeginVerifyAck. The SONET/SDH flags are:
0x0002DCCS: the Test message travels over the Section/RS data communications channel using HDLC framing.0x0004DCCL: the same over the Line/MS DCC.0x0008J0-trace: Section/RS trace correlation.0x0040J1-trace: STS SPE/HOVC path trace correlation.0x0080J2-trace: VT SPE/LOVC path trace correlation.
The values 0x0001, 0x0010 and 0x0020 are reserved.
The neighbors first agree how to pass the test note: write it on a special strip of the cable, or read the name tag already printed on the cable.
- 0x0002DCCS — Test message over the Section/RS DCC.
- 0x0004DCCL — Test message over the Line/MS DCC.
- 0x0008J0-trace — Section/RS trace correlation.
- 0x0040J1-trace — STS SPE/HOVC path trace correlation.
- 0x0080J2-trace — VT SPE/LOVC path trace correlation.
§3.1Trace Correlation for Verification
For the DCC options, the Test message is sent over the data link exactly as defined in RFC 4204. The trace options work differently. With J0, J1 or J2 trace the Test message is not carried in the overhead bytes. It is sent over the control channel and then correlated for consistency with the pattern received in-band in the J0, J1 or J2 byte.
To make this work, the transmitting node must provide the mapping between the in-band pattern and the Test message. That is done with a new TRACE object. The Test message for these modes is the common header, a LOCAL_INTERFACE_ID, a VERIFY_ID and a TRACE object.
Instead of writing the test note on the cable, the neighbor says over the phone "look for the name tag ABC on cable 3". The other side then checks the cable really carries that tag.
§4The TRACE Object and Trace Messages
Trace monitoring lets a node use SONET/SDH capabilities. It can ask a neighbor to monitor a link for a given pattern with TraceMonitor, and the neighbor MUST report a mismatch. It can ask for the current trace value with TraceReq, and it can ask the neighbor to send a specific trace with InsertTrace.
The TRACE object (Class 21, C-Type 1) holds a 16-bit Trace Type, a 16-bit Trace Length and the Trace Message, padded with zeros to a 32-bit boundary. The padding is not counted in Trace Length. Trace Type values are 1 to 6: SONET Section Trace (J0), SONET Path Trace (J1), SONET Path Trace (J2), and the SDH equivalents. The object is non-negotiable.
The memo defines eleven messages, numbered 21 to 31, listed with the diagram below. A TraceMonitorNack uses a new ERROR_CODE C-Type 3 (TRACE_ERROR) with bits 0x01 Unsupported Trace Type and 0x02 Invalid Trace Message. A TraceMonitorNack for a mismatched value means the TraceMismatch message SHOULD NOT be sent as a result.
The messages are questions and answers about name tags on the cable: "watch for this tag", "what tag do you see?", "please put this tag on", and "the tag is wrong".
- 21TraceMonitor — Asks the neighbor to monitor a link for a trace value.
- 22TraceMonitorAck — Request processed, no mismatch.
- 23TraceMonitorNack — Request not supported or the value was wrong.
- 24TraceMismatch — Reports a trace mismatch on a data link.
- 25TraceMismatchAck — Acknowledges TraceMismatch.
- 26TraceReq — Asks for the current trace value of a link.
- 27TraceReport — Returns the requested trace value.
- 28TraceReqNack — Refuses a TraceReq with an error code.
- 29InsertTrace — Asks the neighbor to send a specific trace on a link.
- 30InsertTraceAck — InsertTrace processed correctly.
- 31InsertTraceNack — InsertTrace not processed correctly.
§4.1How the Trace Messages Behave
A node that gets a TraceMonitor request MUST answer with TraceMonitorAck or TraceMonitorNack. The Ack means the TRACE objects were processed correctly, with no trace mismatch. Later, if the in-band trace stops matching, the monitoring node reports it with a TraceMismatch message listing the affected local Interface_Ids, and the receiver MUST respond with TraceMismatchAck. One message can report several links.
A TraceReq names an interface and a Trace Type, and the reply is a TraceReport that MUST include a TRACE object for that data link, or a TraceReqNack with an ERROR_CODE. InsertTrace asks the remote node to send a trace message over a data link, and it assumes the remote knows the mapping between local and remote Interface_Ids. Once InsertTraceAck is received, TraceMismatch is used to flag a mismatch. All of these go over the control channel.
If the name tag on the cable ever changes to the wrong one, a neighbor calls the other and says "cable 3 has the wrong tag", and the other must reply that he heard.
§5Security and IANA
LMP message security uses IPsec as described in RFC 4204. RFC 4207 adds no other new security considerations.
For IANA, the memo assigns LMP Message Types 21 to 31 to the eleven trace messages, TRACE as Object Class 21 (C-Type 1) and TRACE REQ as Object Class 22 (C-Type 1).
The new messages are kept safe the same way as the old LMP messages, and the memo asks for new numbers for them in the official list.