MPLS-TP Still on the Brink

MPLS-TP began as an unprecedented effort by two major standards bodies -- IETF and ITU-T -- to set aside parochial differences and develop a common standard for the necessary OAM tools to enable MPLS meet transport network requirements. The goal was to move quickly and to preserve compatibility with existing MPLS deployments, but in recent months, this ambitious effort has run into the kind of roadblocks that have slowed previous efforts at producing telecom standards more quickly.

For the next few months, Light Reading will be hosting an MPLS-TP Briefing Center, a space devoted to exploring the industry activity around this still-developing standard.

The original thinking that led to MPLS-TP was that a simpler version of MPLS was needed in the transport portion of the network, and while some vendors had designed their owned versions, a standard was needed. Here's a look at coverage of those early efforts at getting MPLS-TP off the ground, back when the technology was duking it out against Provider Backbone Transport (PBT, later PBB-TE):

The thinking was that end-to-end MPLS would simplify metro Ethernet networks and optimize them for transport, enabling more affordable end-to-end MPLS deployments, streamlining operational models and allowing consolidation and simplification of network topologies.

Of course, vendors quickly got involved in developing MPLS-TP right alongside the standards effort. Here's a look at some of their activity:

By mid-2010, MPLS-TP was supposed to be substantially complete as a standard, according to a very ambitious original plan. But delays, typical of the standards process, led some carriers and their vendors to push ahead with an earlier version of MPLS OAM, T-MPLS, as a practical way of meeting immediate needs. Here's a look at those complications, which led an ITU study group to, once again, create a split standard.

For more detail on how MPLS-TP works and its advantages, here's a Heavy Reading white paper:

Light Reading will provide ongoing coverage of both the standards debate and the ongoing deployment of MPLS-TP, so check back at this briefing center.

— Carol Wilson, Chief Editor, Events, Light Reading

Huub_van_Helvoort 12/5/2012 | 5:03:32 PM
re: MPLS-TP Still on the Brink

The opening sentence: "MPLS-TP began as an unprecedented effort by two major standards bodies" is incorrect. OAM for Ethernet was developped in close cooperation between IEEE and ITU-T, this combined effort resulted in IEEE 802.1ag and G.8013/Y.1731. This effort took two years. 

Also the sentence: "while some vendors had designed their owned versions, a standard was needed" is incorrect. The architecture of for the application of MPLS in a transport network was developed in 2005/2006 by the ITU-T, based on this architecture the OAM for this architecture was developed, and described in G.8114 (T-MPLS OAM). G.8114 was ready for approval in January 2008. It was also implemented by several vendors, and has been deployed in operational networks.

The sentence: "By mid-2010, MPLS-TP was supposed to be substantially complete as a standard" is incorrect. According to the Joint (ITU-T and IETF) Working Team MPLS-TP OAM should have been available in September 2009. The only MPLS-TP OAM described at that time was based on G.8013.Y.1731 OAM and available as draft in ITU-T: G,tpoam and the IETF: draft-bhh-mpls-tp-oam-y1731 to be used in a Packet Transport Network (PTN).

Your list of activities does not mention successfull interop testing of G.tpoam by EANTC in 09/2009, 02/2010 and 09/2010, nor does it mention "China Mobile, the successful organisation of the world's largest equipment interoperability testing PTN".

I don't agree with your statement "delays, typical of the standards process". The ITU-T uses the Advanced Approval Process (AAP) to shorten the standards process. The use of the Traditional Approval Process (TAP) is exceptional, voting by Member States on the approval is very exceptional in ITU-T.

Your statement "complications, which led an ITU study group to, once again, create a split standard" is incorrect, there may be (once again) a split in opinions, but there are many ITU-T members that have no objection to describe two MPLS-TP OAM toolsets, each for its own application. It is the IETF that does not accept this compromise.

Additional information can be found on the ITU-T newslog:

MPLS-TP The Facts
cast doubt on ‘jeopardize’ Internet statement
ITU satisfies market demand for carrier class MPLS standard - Network operators given cost-effective management solution

Sign In