New Answers for Bringing Scale to the Carrier Ethernet Metro Edge – Part 1

Aug 27, 2013 by Brian Van Voorhis

Brian Van Voorhis of OvertureMany service providers are searching for new ways to scale up their service capacity and speed service delivery as they transition their edge and aggregation network from a legacy non-MEF network to one that is Carrier Ethernet CE 2.0 compliant. In this blog series we outline an optimized architecture vision that helps service providers migrate from a non-MEF edge/aggregation network to a MEF-compliant network.   The factors we consider are a) the embedded base of bridging elements in the edge /aggregation network b) the establishment of MPLS in the core and c) how software defined networking (SDN) and network functions virtualization (NFV) concepts potentially alter the landscape in planning a solution. 

The new architecture proposed in the final part of the blog creates a MEF-compliant network, reduces costs, maximizes operational efficiencies and introduces new revenue-generating services on a scale never before possible.

The blog discussion consists of three parts

Part 1 – Problems? What Problems?

In recent years, we have seen a debate about the future of the edge network.  The debaters fall into one of two groups which propose solutions that a) extend MPLS pseudowires to the edge of the network or b) build a Carrier Ethernet (CE) 2.0 aggregation/edge network to provide on-ramps to the MPLS core.   The opposing camps often provide complex arguments for their positions.  Many times these discussions fail to start with the most primary question, “what problems are we trying to solve?”

Let’s examine a real-world carrier network example.  The network diagram below illustrates a number of the problems facing this particular service provider. Their network has evolved to a complex mix of technologies that are not able to deliver a MEF-compliant service offering.  As you can see the network consists of an MPLS core with a legacy, layer 2 STP aggregation network. 

Carrier Ethernet Scaling Challenge at the Metro Edge

As a result of this network architecture, our customer has shared five key problems they need to solve:

  1. MAC Address Scaling.   MAC scaling issues largely occur in legacy L2 networks because they rely on bridged networks to deliver E-Line services.  MEF CE 2.0 E-Line networks, being transparent to MAC addresses, largely eliminate the issues with table size, learning rate and flooding rate limitations found in legacy networks.
  2. Supporting Wholesale Ethernet Customers.  It is nearly impossible to provide MEF E-access service across a legacy non-MEF network.
    1. The Ethernet service providers require MEF service assurance metrics (delay PM and loss PM) to the CPE to assure the layer 2 service performance.
    2. Crossing operator networks require strong capabilities for stacked VLAN manipulations and mapping between different Class of Service domains.
    3. The Ethernet access operator needs to monitor and report on the connectivity status between their endpoints and the hand off to the Ethernet service provider
    4. In the case of remote handoffs, the Ethernet access operator needs to know the connectivity status and health to the remote hand-off site.
  3. L2 and L3 Services.  Some business customers buy layer 2 services and some buy layer 3. This drives both layer 2 and layer 3 CPE.  Service changes cause truck rolls and redeployment of CPE equipment.
  4. Fiber/Copper Media.  End user sites are connected using a mix of fiber and legacy copper technologies.  The end user experience and SLA must be consistent for service delivered over fiber and copper.
  5. Service Assurance.  The goal is to utilize service assurance OAM models that 1) directly correspond to the delivered service, 2) are industry standard and 3) are easy to extend across wholesale interfaces with minimum complexity.

These are commonly identified problems faced by operators at the metro edge.  If you can identify with these in your own experience, then you should be interested in Part 2 where we explore some options for addressing them.

About the Author

Brian Van Voorhis, is Senior Product Manager for the 6500 Service Delivery Platform and Ensemble Network Controller at Overture Networks.  He is also a Carrier Ethernet Certified Professional (MEF CECP).  With extensive experience in edge an d aggregation networks Brian has successfully launched and managed high profile product development initiatives during his career, leveraging Carrier Ethernet, Active Ethernet, GPON, DSL, and SONET/TDM technologies.   Brian has a BS Degree in Electrical Engineering from University of Cincinnati and an MS in Electrical Engineering from Florida Atlantic University.

Comments

Post new comment

The content of this field is kept private and will not be shown publicly.
By submitting this form, you accept the Mollom privacy policy.
To prevent automated spam submissions leave this field empty.