GMPLS Ethernet UNI: Generalized MPLS (GMPLS) Support for Metro Ethernet Forum and G.8011 User Network Interface (UNI)
How a customer's edge node asks a provider network for an Ethernet connection using GMPLS signaling.
Berger, L. and D. Fedyk, "Generalized MPLS (GMPLS) Support for Metro Ethernet Forum and G.8011 User Network Interface (UNI)", RFC 6005, DOI 10.17487/RFC6005, October 2010, <https://www.rfc-editor.org/info/rfc6005>.
@misc{rfc6005,
series = {Request for Comments},
number = 6005,
howpublished = {RFC 6005},
publisher = {RFC Editor},
doi = {10.17487/RFC6005},
url = {https://www.rfc-editor.org/info/rfc6005},
author = {L. Berger and D. Fedyk},
title = {{Generalized MPLS (GMPLS) Support for Metro Ethernet Forum and G.8011 User Network Interface (UNI)}},
pagetotal = 10,
year = 2010,
month = oct,
} RFC text
Contents
Internet Engineering Task Force (IETF) L. Berger
Request for Comments: 6005 LabN
Category: Standards Track D. Fedyk
ISSN: 2070-1721 Alcatel-Lucent
October 2010
Generalized MPLS (GMPLS) Support for Metro Ethernet Forum
and G.8011 User Network Interface (UNI)
Abstract
This document describes a method for controlling two specific types
of Ethernet switching via a GMPLS-based User Network Interface (UNI).
This document supports the types of switching required by the
Ethernet services that have been defined in the context of the Metro
Ethernet Forum (MEF) and International Telecommunication Union (ITU)
G.8011. This document is the UNI companion to "Generalized MPLS
(GMPLS) Support for Metro Ethernet Forum and G.8011 Ethernet Service
Switching". This document does not define or limit the underlying
intra-domain or Internal NNI (I-NNI) technology used to support the
UNI.
Status of This Memo
This is an Internet Standards Track document.
This document is a product of the Internet Engineering Task Force
(IETF). It represents the consensus of the IETF community. It has
received public review and has been approved for publication by the
Internet Engineering Steering Group (IESG). Further information on
Internet Standards is available in Section 2 of RFC 5741.
Information about the current status of this document, any errata,
and how to provide feedback on it may be obtained at
http://www.rfc-editor.org/info/rfc6005.
Copyright Notice
Copyright (c) 2010 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
Table of Contents
1. Introduction ....................................................3
1.1. Overview ...................................................4
1.2. Conventions Used in This Document ..........................5
2. Common Signaling Support ........................................5
2.1. UNI Addressing .............................................5
2.2. Ethernet Endpoint (UNI) Identification .....................6
2.2.1. Address Resolution ..................................6
2.3. Connection Identification ..................................7
3. EPL Service .....................................................7
4. EVPL Service ....................................................7
4.1. Egress VLAN ID Control and VLAN ID Preservation ............7
5. IANA Considerations .............................................8
5.1. Error Value: Routing Problem/Unknown Endpoint ..............8
6. Security Considerations .........................................8
7. References ......................................................8
7.1. Normative References .......................................8
7.2. Informative References .....................................9
Acknowledgments ....................................................9
1. Introduction
[MEF6] and [G.8011] provide parallel frameworks for defining network-
oriented characteristics of Ethernet services in transport networks.
The framework discusses general Ethernet connection characteristics,
Ethernet User Network Interfaces (UNIs), and Ethernet Network-Network
Interfaces (NNIs). Within this framework, [G.8011.1] defines the
Ethernet Private Line (EPL) service and [G.8011.2] defines the
Ethernet Virtual Private Line (EVPL) service. [MEF6] covers both
service types. [MEF10.1] defines service parameters and [MEF11]
provides UNI requirements and framework.
This document provides a method for GMPLS-based control of Label
Switched Paths (LSPs) that support the transport services defined in
the above documents at the UNI network reference points. This
document does not define or limit the underlying intra-domain or
Internal NNI (I-NNI) technology used to support the UNI. This
document makes use of the GMPLS extensions defined in [RFC6004] and
[RFC6002].
The scope of this document covers Ethernet UNI applications, and it
is intended to be consistent with the GMPLS overlay model presented
in [RFC4208] and aligned with GMPLS Core Network signaling. The
scope and reference model used in this document are represented in
Figure 1, which is based on Figure 1 of [RFC4208].
Figure 1 shows two core networks, each containing two core nodes.
The core nodes are labeled 'CN'. Connected to each CN is an edge
node. The edge nodes are labeled 'EN'. Each EN supports Ethernet
Networks and use Ethernet services provided by the core nodes via a
UNI. Two services are represented: one EPL and one EVPL type
service. Signaling within the core network is out of scope of this
document and may include any technology that supports overlay UNI
services. The UNI function in the edge node can be referred to as
the UNI client, or UNI-C, and in the CN as UNI network, or UNI-N.
Ethernet Ethernet
Network +----------+ +-----------+ Network
+---------+ | | | | +---------+
| +----+ | | +-----+ | | +-----+ | | +----+ |
------+ | | EPL | | | | | | | | EPL | | +------
------+ EN +-+-----+--+ CN +---------+ CN +--+-----+-+ EN +------
| | | | +--+--| +---+ | | +--+-----+-+ | |
| +----+ | | | +--+--+ | | | +--+--+ | | +----+ |
| | | | | | | | | | | |
+---------+ | | | | | | | | +---------+
| | | | | | | |
+---------+ | | | | | | | | +---------+
| | | | +--+--+ | | | +--+--+ | | |
| +----+ | | | | | | +-----+ | | | +----+ |
------+ +-+--+ | | CN +---------+ CN | | | | +------
------+ EN +-+-----+--+ | | | | +--+-----+-+ EN +------
| | | |EVPL | +-----+ | | +-----+ |EVPL | | | |
| +----+ | | | | | | +----+ |
| | +----------+ |-----------+ | |
+---------+ Core Network(s) +---------+
Ethernet UNI UNI Ethernet
Network <-----> <-----> Network
Scope of This Document
Legend: EN - Edge Node
CN - Core Node
Figure 1: Ethernet UNI Reference Model
1.1. Overview
This document uses a common approach to supporting the switching
implied by the Ethernet services defined in [MEF6], [G.8011.1], and
[G.8011.2]. The approach builds on standard GMPLS mechanisms to
deliver the required control capabilities. This document reuses the
GMPLS mechanisms specified in [RFC6004], [RFC4208], and [RFC4974].
Support for Point-to-Point (P2P) and Multipoint-to-Multipoint (MP2MP)
service is required by [G.8011] and [MEF11]. P2P service delivery
support is based on the GMPLS support for Ethernet services covered
in [RFC6004]. As with [RFC6004], the definition of support for MP2MP
service is left for future study and is not addressed in this
document.
[MEF11] defines multiple types of control for UNI Ethernet services.
In MEF UNI Type 1, services are configured manually. In MEF UNI Type
2, services may be configured manually or via a link management
interface. In MEF UNI Type 3, services may be established and
managed via a signaling interface. As with [RFC6004], this document
is aimed at supporting the MEF UNI Type 3 mode of operation (and not
MEF UNI Types 1 and 2). As mentioned above, this document is limited
to covering UNI-specific topics.
Common procedures used to signal Ethernet connections are described
in Section 2 of this document. Procedures related to signaling
switching in support of EPL services are described in Section 3.
Procedures related to signaling switching in support of EVPL services
are described in Section 4.
1.2. Conventions Used in This Document
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].
2. Common Signaling Support
This section describes the common mechanisms for supporting a UNI
reference point for LSPs that provide the Ethernet Services described
in [RFC6004].
Except as specifically modified in this document, the procedures
related to the processing of Resource ReSerVation Protocol (RSVP)
objects is not modified by this document. The relevant procedures in
existing documents, notably [RFC6002], [RFC6004], [RFC4208], and
[RFC4974], MUST be followed in all cases not explicitly described in
this document.
2.1. UNI Addressing
LSPs providing Ethernet connections controlled via the mechanisms
defined in this document MUST use the addressing and other procedures
defined in [RFC4208]. Of note, this includes the use of the egress
edge node's IP address in the endpoint address field in the SESSION
object.
One issue that presents itself with the addressing approach taken in
[RFC4208] is that an ingress edge node may not receive the egress
edge node's IP address as part of the management, or other, request
that results in the initiation of a new Ethernet connection. This
case is covered as described in Section 7.2 of [RFC4974] and modified
below in Section 2.2.1.
2.2. Ethernet Endpoint (UNI) Identification
UNI identification, except as noted below, MUST follow Ethernet
endpoint (UNI) identification as defined in [RFC6004]. There is one
additional case that is covered in this document where the scope of
the Ethernet endpoint identifier is relevant beyond the typical case
of just ingress and egress nodes.
2.2.1. Address Resolution
At the UNI reference point, it is possible for the ingress edge node
to not have the egress edge node's IP address when initiating an LSP.
This presents an issue as the egress edge node's IP address is
carried in the SESSION object. This case is handled leveraging the
approach described in Section 7.2 of [RFC4974] to address call ID
assignment by the first core node.
When an edge node (the UNI-C) initiates an LSP and it has the egress
Ethernet endpoint identifier, but does not have its IP address, the
edge node MUST create a Notify message as described in [RFC4974].
The Notify message MUST include the CALL_ATTRIBUTES object with the
Endpoint ID TLV defined [RFC6004]. The tunnel endpoint address field
of the SESSION object in the Notify message MUST be set to zero (0).
The message MUST be addressed and sent to an address associated with
the first core node.
When a core node, i.e., the node providing the network side of the
UNI (the UNI-N), receives a Notify message with the tunnel endpoint
address field of the SESSION object set to zero, it MUST locate the
Endpoint ID TLV in the CALL_ATTRIBUTES object. If the object or TLV
are not present, the node MUST discard the message. In this case, a
Message ID Acknowledgment MUST NOT be sent for the Notify message.
When the Endpoint ID TLV is located, the node MUST map the Endpoint
ID into an IP address associated with the egress edge node. If the
node is unable to obtain an egress address, it MUST issue an error
response Notify messages according to Section 6.2.2. of [RFC4974].
The Error code and value SHOULD be "Routing Problem/Unknown Endpoint"
(Error code 24, Error value 35).
When the node is able to obtain an egress address, the endpoint
address field of the SESSION object MUST be set to the obtained
address, and the Notify message should be sent according to the
standard processing defined in [RFC4974]. The downstream nodes will
then process the Notify according to standard processing rules.
When the ingress receives the response Notify message, it SHOULD
identify the call based on the Endpoint ID TLV and, when not set to
zero on the corresponding setup Notify message, the short and long
Call IDs. The endpoint address field of the SESSION object carried
in the response Notify message will include the egress's IP address.
This returned address MUST be used in all subsequent messages
associated with the Ethernet connection.
Note that the procedure described in this section MAY be used when
the Call IDs are generated by the initiating UNI or generated by the
first core node.
2.3. Connection Identification
With one exception, UNI signaling for Ethernet connections MUST
follow the Connection Identification procedures defined in [RFC6004].
The exception is that the procedures defined in Section 7.2 of
[RFC4974] MAY be used to provide support for allocation of Call IDs
by the first core node rather than by the initiating edge node.
3. EPL Service
There are no additional UNI-specific requirements for signaling LSPs
supporting Ethernet Private Line (EPL) services. The procedures
defined in [RFC6004], as modified above, MUST be followed when
signaling an LSPs supporting an EPL Service.
4. EVPL Service
There is one additional UNI-specific requirement for signaling LSPs
supporting an EVPL type service as described in Section 4.1. Except
as modified above and by this section, the procedures defined in
[RFC6004] MUST be followed when signaling an EVPL Service.
4.1. Egress VLAN ID Control and VLAN ID Preservation
Per [MEF6], the mapping of the single VLAN ID used at the ingress UNI
to a different VLAN ID at the egress UNI is allowed for EVPL services
that do not support both bundling and VLAN ID preservation. Such a
mapping MUST be requested and signaled based on the explicit label
control mechanism defined in [RFC4208], and not the mechanisms
defined in [RFC6004].
As is the case in [RFC6004], when the explicit label control
mechanism is not used VLAN IDs MUST be preserved, i.e., not modified,
across the LSP.
5. IANA Considerations
IANA has assigned new values for namespaces defined in this document
and summarized in this section.
5.1. Error Value: Routing Problem/Unknown Endpoint
IANA has made the following assignment in the "Error Codes and
Globally-Defined Error Value Sub-Codes" section of the "RSVP
Parameters" registry located at http://www.iana.org:
Error Code Meaning
24 Routing Problem [RFC3209]
Under "This Error Code has the following globally-defined Error
Value sub-codes:"
35 = Unknown Endpoint [RFC6005]
6. Security Considerations
This document makes use of the mechanisms defined in [RFC6004] and
[RFC4974]. It does not in itself change the security models offered
in each. (Note that the address resolution discussed in Section 2.2
above, parallels the replacement of information that occurs per
Section 7.2 of [RFC4974].) See [RFC6004] and [RFC4974] for the
security considerations that are relevant to and introduced by the
base mechanisms used by this document.
7. References
7.1. Normative References
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC3209] Awduche, D., Berger, L., Gan, D., Li, T., Srinivasan, V.,
and G. Swallow, "RSVP-TE: Extensions to RSVP for LSP
Tunnels", RFC 3209, December 2001.
[RFC4208] Swallow, G., Drake, J., Ishimatsu, H., and Y. Rekhter,
"Generalized Multiprotocol Label Switching (GMPLS) User-
Network Interface (UNI): Resource ReserVation Protocol-
Traffic Engineering (RSVP-TE) Support for the Overlay
Model", RFC 4208, October 2005.
[RFC4974] Papadimitriou, D. and A. Farrel, "Generalized MPLS (GMPLS)
RSVP-TE Signaling Extensions in Support of Calls", RFC
4974, August 2007.
[RFC6002] Berger, L. and D. Fedyk, "Generalized MPLS (GMPLS) Data
Channel Switching Capable (DCSC) and Channel Set Label
Extensions", RFC 6002, October 2010.
[RFC6004] Berger, L. and D. Fedyk, "Generalized MPLS (GMPLS) Support
for Metro Ethernet Forum and G.8011 Ethernet Service
Switching", RFC 6004, October 2010.
7.2. Informative References
[G.8011] ITU-T G.8011/Y.1307, "Ethernet over Transport Ethernet
services framework", August 2004.
[G.8011.1] ITU-T G.G.8011.1/Y.1307.1, "Ethernet private line
service", August 2004.
[G.8011.2] ITU-T G.8011.2/Y.1307.2, "Ethernet virtual private line
service", September 2005.
[MEF6] The Metro Ethernet Forum, "Ethernet Services Definitions -
Phase I", MEF 6, June 2004.
[MEF10.1] The Metro Ethernet Forum, "Ethernet Services Attributes
Phase 2", MEF 10.1, November 2006.
[MEF11] The Metro Ethernet Forum , "User Network Interface (UNI)
Requirements and Framework", MEF 11, November 2004.
Acknowledgments
Dimitri Papadimitriou provided substantial textual contributions to
this document and coauthored earlier versions of this document.
The authors would like to thank Evelyne Roch and Stephen Shew for
their valuable comments.
Authors' Addresses
Lou Berger
LabN Consulting, L.L.C.
Phone: +1-301-468-9228
EMail: lberger@labn.net
Don Fedyk
Alcatel-Lucent
Groton, MA 01450
Phone: +1-978-467-5645
EMail: donald.fedyk@alcatel-lucent.com
AI ELI5
On this page
§1Scope and Reference Model
The Metro Ethernet Forum (MEF) and ITU-T G.8011 define Ethernet services for transport networks, including the Ethernet Private Line (EPL) and Ethernet Virtual Private Line (EVPL) services. RFC 6005 describes how GMPLS can control Label Switched Paths (LSPs) that carry those services at the User Network Interface (UNI).
The reference model has edge nodes (EN) that connect to Ethernet networks and core nodes (CN) in the provider network. The UNI function in an edge node is the UNI-C (client) and the one in a core node is the UNI-N (network). What happens inside the core network is out of scope: the RFC does not define or limit the technology used there.
It builds on RFC 4208 (the GMPLS overlay model), RFC 6004 (Ethernet service switching) and RFC 6002. It is the UNI companion to RFC 6004.
A customer's building plugs into the phone company's network at a front door called the UNI. This RFC only covers what happens at that front door, not inside the phone company.
§1.1Which Services and Control Modes Are Covered
The RFC supports point-to-point (P2P) service. Multipoint-to-multipoint (MP2MP) service is required by G.8011 and MEF 11, but, as in RFC 6004, it is left for future study and not addressed here.
MEF 11 defines three types of UNI control. In Type 1 services are configured manually. In Type 2 they may be configured manually or through a link management interface. In Type 3 they can be established and managed through a signaling interface. This document targets Type 3 only.
Apart from the changes described in the RFC, existing procedures for RSVP objects are not modified, and those in RFC 6002, RFC 6004, RFC 4208 and RFC 4974 MUST be followed.
There are three ways to order a connection: by hand, by hand with a helper tool, or by sending messages automatically. This RFC is only about the automatic way.
§2.1UNI Addressing and Endpoint Identification
LSPs controlled under this RFC MUST use the addressing and procedures of RFC 4208. This includes putting the egress edge node's IP address in the endpoint address field of the SESSION object.
The catch is that the ingress edge node may not have that address. The request that starts a connection might only name the far Ethernet endpoint. Ethernet endpoint (UNI) identification otherwise follows RFC 6004, with one extra case covered by the address resolution procedure.
To mail a letter you need the street address. Sometimes the customer only knows the name of the person to reach, so the address has to be looked up first.
§2.2Address Resolution with Notify
When an edge node has the egress Ethernet endpoint identifier but not its IP address, it MUST send a Notify message as described in RFC 4974. The message MUST carry a CALL_ATTRIBUTES object with the Endpoint ID TLV from RFC 6004, and the tunnel endpoint address of the SESSION object MUST be set to zero (0). It is sent to an address of the first core node.
A core node (UNI-N) that receives such a Notify MUST find the Endpoint ID TLV. If the object or TLV is missing, it MUST discard the message and MUST NOT send a Message ID Acknowledgment. Otherwise it MUST map the endpoint ID to an IP address of the egress edge node and set the endpoint address field of the SESSION object to it, then continue standard RFC 4974 processing.
If no address can be found, the node MUST send an error Notify (Section 6.2.2 of RFC 4974). The error code and value SHOULD be Routing Problem/Unknown Endpoint (error code 24, error value 35).
When the ingress receives the response Notify, it SHOULD identify the call from the Endpoint ID TLV and, if they were not zero in the setup Notify, the short and long Call IDs. The returned address MUST be used in all later messages for that Ethernet connection.
The customer asks the carrier, "Who is Bob's mailbox?" The carrier looks it up and answers with the address. If the carrier cannot find Bob, it sends back an error that says so.
- Notify, SESSION address 0Carries CALL_ATTRIBUTES with the Endpoint ID TLV.
- Notify, SESSION has egress IPOr an error: Routing Problem/Unknown Endpoint (24, 35).
§2.3Connection Identification and Call IDs
UNI signaling for Ethernet connections MUST follow the connection identification procedures of RFC 6004, with one exception. The procedures of Section 7.2 of RFC 4974 MAY be used so that the first core node, rather than the initiating edge node, allocates the Call IDs.
The address resolution procedure above MAY be used whether the Call IDs are generated by the initiating UNI or by the first core node.
Each connection needs a ticket number. Usually the customer picks it, but this RFC lets the carrier hand out the number instead.
§3EPL and EVPL Services
For EPL, there are no additional UNI-specific requirements. The procedures of RFC 6004, as modified by the sections above, MUST be followed.
For EVPL there is one extra requirement. MEF 6 allows the single VLAN ID used at the ingress UNI to be mapped to a different VLAN ID at the egress UNI, for EVPL services that do not support both bundling and VLAN ID preservation. Such a mapping MUST be requested and signaled with the explicit label control mechanism of RFC 4208, and not the mechanisms of RFC 6004. When explicit label control is not used, VLAN IDs MUST be preserved, meaning they are not modified across the LSP.
A private line needs no special extras. A virtual line may relabel the customer's traffic tag at the far end, but only if the customer asks for that in a specific way. Otherwise the tag stays the same.
§4IANA Assignment
IANA assigned error value 35, Unknown Endpoint, under error code 24 (Routing Problem) in the "Error Codes and Globally-Defined Error Value Sub-Codes" part of the RSVP Parameters registry. Error code 24 itself comes from RFC 3209.
The official list of error messages got one new line: "Unknown Endpoint".
§5Security Considerations
The RFC uses the mechanisms of RFC 6004 and RFC 4974 and does not itself change the security models of either. It notes that the address resolution step parallels the replacement of information that occurs under Section 7.2 of RFC 4974. Readers are pointed to those two RFCs for the relevant security considerations.
This RFC adds no new locks or keys. It relies on the ones from the RFCs it builds on.