The Multiprotocol Label Switching - Transport Profile (MPLS-TP) is a packet transport technology that leverages benefits from the existing MPLS networking infrastructure, and improves the efficiency and effectiveness of packet transport networks, while maintaining the mandatory Operation Administration and Maintenance (OAM) capabilities of legacy SONET/SDH networks.
This technical note provides information on MPLS-TP technologies and implementation in Orckit-Corrigent's Packet Transport Network (PTN) solutions.
MPLS-TP is based on a subset of MPLS technologies with additional transport functionalities, as in traditional transport networks, making it a reliable, scalable, and cost optimized packet-based transport technology.
MPLS-TP introduces additional transport functionalities such as comprehensive OAM capabilities, survivability, data-plane/control-plane separation, and static provisioning of bidirectional services. This is in addition to well-accepted MPLS functionalities such as Quality of Service (QoS), scalability, traffic engineering and Layer 2 packet forwarding. The result is the ability to provide network operators with full control over their packet networks.
MPLS-TP includes the following main functionalities defined by Internet Engineer Task Force (IETF) and International Telecommunications Union Telecommunications Standardization Sector (ITU-T) standardization bodies
- Bidirectional Label Switch Path (LSP)
- Transmit and receive traffic following the same path throughout the network
- Enhanced OAM tools
- Continuity Check (CC) for fast failure detection
- Alarm Indication Signal (AIS) for fault isolation
- Remote Defect Indication (RDI) for fault isolation
- Loopback (LB), similar to IP ping, for basic maintenance
- Loss/delay measurement for detection of performance degradation
- End-to-end Protection
- 1:1 bidirectional end-to-end protection scheme triggered by OAM
- Sub 50ms linear protection
- Pseudo wire (PW) and LSP protection
- Control plane options
- Separation from the data plane
- Static or dynamic LSP configuration
MPLS-TP in the standardization bodies
The IETF and the ITU-T teams agreed to work together on the design of MPLS-TP, in order to bring transport network requirements to the existing MPLS technology.
The Joint Working Group (JWT) focuses on the following main categories
- General requirements
- Network management
The JWT started their work in March 2008. The first draft was published in July 2008 and the final agreement is expected at the end of 2011.
The need for MPLS-TP
The increasing growth of packet-based traffic is driving the need for evolved transport networks. The next generation transport technologies should leverage the same benefits of legacy transport in terms of end-to-end determinism, OAM, high availability, high reliability, and easy to use management.
Figure #1: MPLS-TP evolution
MPLS-TP OAM provides the same OAM concepts and methods which are available in legacy transport networks, including
- Fast failure detection
- Alarm suppression
- Remote defect indication
- Protection switching
OAM packets are carried on Generic Associated channel (G-Ach).
OAM and data packets are carried on the same path, therefore enabling simpler and faster monitoring of the PW and LSP layers.
MPLS-TP OAM introduces the functional components, Maintenance End Point (MEP) and Maintenance Intermediate Point (MIP), which enable running OAM packets between two end points, such as:
1. Continuity checks (CC) messages allowing fast detection of lost of connectivity, as well as connection mis-configuration.
The CC messages are sent periodically by MEP and monitored for Loss Of Continuity (LOC) by each MEP/MIP.
The transmission interval can be set to a minimum of 3.3 ms, allowing for very fast-failure detection.
Figure #2: CCM Flow
2. Delay and Loss Measurements (DM/LM) allowing the detection of performance degradations.
3. Loopback (LB) messages allowing on-demand bidirectional diagnostic test for connectivity check and failure localization.
4. Alarms suppress enabling fault localization while avoiding unnecessary alarms propagation. AIS and RDI are sent to the remote sides in case of LOC detection.
MPLS-TP provides similar protection switching mechanisms which are used in legacy SONET/SDH transport networks, including
- End to end linear protection for LSP, PW and Multi Segment PW (MS-PW) layers
- Protection State Coordination (PSC) mechanism to synchronize the both ends on protection status and commands
- Sub-50 ms protection in case of failure in the data plane (relying on OAM messages transmission at 3.3 ms rate)
- Manual protection commands e.g. manual switch, lockout, clear
- Provides architecture of 1+1 and 1:1 protection schemes
- Protection switching is triggered upon Signal Fail (i.e. LOC) or Signal Degrade (SD) or manual commands
All these mechanisms provide enhanced LSP/PW/MS-PW end- to-end protection, ensuring network reliability and high availability.
MPLS-TP control plane and management
The control plane in MPLS-TP is the mechanism used for provisioning LSPs/PWs dynamically over the packet network. The control plane is optional and separated from the data plane. A failure of the control plane will not create any kind of data plane failure.
When no control plane is used, operators can set up LSPs and PWs statically using a Network Management System (NMS), similar to the way it is done in legacy transport networks, without IP or routing protocols.
In the dynamic approach, a control plane is used. LSPs and PWs creation is done by Generalize MPLS (GMPLS) and Targeted Label Distribution Protocol (T-LDP).
Orckit-Corrigent MPLS-TP value proposition
IP/MPLS is field-proven and a de-facto standard in the carrier Ethernet market, while MPLS-TP is an emerging, and yet to be ratified standard that brings additional benefits (OAM, protection, simplicity) to packet-based networks. The uncertainty of MPLS-TP complete definition is a risk for service providers that wish to benefit from the technology capabilities, but are looking for standard and interoperable solutions.
Orckit-Corrigent's CM-4000 PTN solutions offer dual stacking of IP/MPLS and MPLS-TP on the same hardware and software!
The dual stack IP/MPLS and MPLS-TP solution allows the service provider
- A delay in the need to make a decision regarding MPLS-TP until its acceptance and applicability is clear
- Full network flexibility to select IP/MPLS and/or MPLS-TP per LSP per application per region
- Inherent interoperability between the IP/MPLS domain and MPLS-TP domain
- Broader gateway between existing IP-MPLS deployments and emerging MPLS-TP deployments
Figure #3: Orckit-Corrigent IP/MPLS and MPLS-TP interworking
Together with the dual stacking of IP/MPLS and MPLS-TP, Orckit-Corrigent CM-4000 offers a centralized NMS (CM-View) with a Path Computation Element (PCE), Traffic Engineering Database (TED) and Call Admission Control (CAC). This solution enables
- Automatic provisioning of LSPs considering Traffic Engineering (TE) properties such as Class of Service (CoS) and bandwidth
- Understanding the relationship between network resources and LSPs
- Creation of manual or automatic explicit route
- Support of a disjointed path for LSP/PW protection
Migration to MPLS-TP
Service providers who wish to deploy a new packet-based transport network based on MPLS-TP technology are facing a dilemma of which technology to use as a stepping stone until MPLS-TP will be ratified. Such intermediate step should take into account the complexity and the cost of migrating at the end to MPLS-TP. There are 2 options for such intermediate step:
- T-MPLS technology
- IP/MPLS Technology
In the first option, the T-MPLS which is a non-standard technology is positioned as a stepping stone to MPLS-TP. The assumption is that since T-MPLS is the basic of MPLS-TP, the migration in the future will only be a software upgrade, but since the MPLS-TP standard still has a long way to go, there is a risk that service providers will end up with a complicated and expensive migration.
In the second approach, the service providers can choose the IP/MPLS technology as a step towards the MPLS-TP, the IP/MPLS is field-proven and a de-facto standard unlike T-MPLS.
In such approach the service providers will eliminate the risk of using a non-standard technology and in addition they will have the ability to migrate only part of their IP/MPLS network to MPLS-TP due to the interworking between the 2 technologies.
Orckit-Corrigent CM-4000 products support both IP/MPLS and MPLS-TP simultaneously making the migration process smooth and cost-efficient enabling service providers to migrate to the standardized MPLS-TP by using a software upgrade only and saving the cost of new hardware.
MPLS-TP brings transport characteristics such as OAM, survivability and management to the MPLS domain, while preserving its flexibility and scalability.
Orckit-Corrigent's CM-4000 PTN solutions support IP/MPLS and MPLS-TP simultaneously, and by doing this enable the service provider a smooth migration from IP/MPLS to MPLS-TP as well as interoperability between the two technologies.
Orckit-Corrigent actively participates in the MPLS-TP standardization efforts and its CM-4000 products already support MPLS-TP. The CM-4000 MPLS and MPLS-TP interoperability was demonstrated at Carrier Ethernet World Congress, Warsaw during September 2010.