CE-ing Carrier Ethernet’s Future
Ethernet has become the ubiquitous device-to-device interconnection, displacing almost all other legacy data transport. If you have to connect anything that uses data, you’re probably using Ethernet to do it. But Ethernet is 40 years old this year. Years of technology advances and innovation have brought options and implementation intricacies that can get the heads spinning on even the most qualified engineers. Interface types, transmission rates and operating characteristics are captured in specifications that are literally thousands of pages long, from IEEE, ITU-T and other standards bodies.
Of course, most communications service providers, both internal and external, just want to plug it in and expect it to work.
Enter the Metro Ethernet Forum (MEF). Since 2001, this body has defined the “how” of setting up Ethernet services from one end of a circuit to the other. Because there are so many ways that Ethernet can be used, the MEF came together to define what are the most useful ways to deploy. Since its inception, the MEF has had strong representation from service providers from around the world. Because of the high percentage of service providers (greater than 50% of its membership), the MEF is the right standards body to define the requirements for Carrier Ethernet networks.
By 2005-2006, the MEF introduced the first generation of equipment certifications. These certifications brought basic organization to the fledgling Carrier Ethernet market. The goal achieved by those first certifications was to establish dependable Ethernet service over any transport technology. (See Figure 1.) So native Ethernet services or Ethernet services over PDH (T1, DS3, E1), Frame Relay/ATM, SONET/SDH or OTN/WDM always perform within the expectations of end-user service level agreements (SLAs).
Figure 1. Ethernet deployment models (graphic provided courtesy of MEF).
MEF certification plays a large role in the growth of Carrier Ethernet services. In fact, last year the share of bandwidth deployed as Ethernet services exceeded the sum total of all other legacy telecom services, making it the dominant technology in the carrier space, according to Vertical Systems Group. During the course of 2012, over 1.2 billion new Ethernet ports were deployed: 400 million wired and 800 million wireless, according to IDC.
A New Suite Spot -- Certifications
In January 2013, the MEF introduced a new suite of certifications for Ethernet equipment manufacturers. This is part of a concerted effort by the Ethernet industry to qualify the next generation of Carrier Ethernet services that should be deployed for the foreseeable future. The new certification suite, labeled Carrier Ethernet 2.0 (or just CE 2.0), strategically tests elements of the MEF standards that have been developed over the last five years.
CE 2.0 brings three major areas of improvement to Carrier Ethernet services that directly impact the end-user’s quality of experience:
Improvement 1: Multiple classes of service (Multi-CoS)
Improvement 2: Enhanced operations, administration and maintenance (OAM)
Improvement 3: Interoperability and E-Access
These improvements are a major consideration of why operations will require CE 2.0 certified equipment and services in the immediate future.
Ethernet services are adversely impacted by oversubscription when bandwidth usage/demand exceeds the bandwidth availability. Oversubscription is employed because the nature of data traffic is burst-y. For any given service between centers, engineers determine the best size of the Ethernet service based on the type and volume of traffic between the centers. Unfortunately, applications tend to grow. The right-size service on Day 1 can quickly become congested as the applications grow in usage over time.
In most of today’s point-to-point Ethernet services, all the data looks the same. One Ethernet frame is not distinguished from another. In times of congestion there is no mechanism to control what kind of traffic gets dropped. In addition, the applications sharing the bandwidth between the centers are not all the same. Their performance and delay requirements may be quite different. If there is no distinction between one Ethernet frame and another in times of congestion, it could be your colleague’s party invitation emails that get bogged down and require retransmission -- or it could be the critical time-of-day stamp on your financial marketing transactions that get dropped.
This issue is addressed by CE 2.0’s certification of Multi-CoS features. The value of assigning a certain class of service to different applications is that it identifies when one application is more important than another. (See Figure 2.) CE 2.0 certifies the minimal set of high, medium and low classes of service so that in times of congestion high class of service traffic is prioritized, or guaranteed with appropriate engineering guidelines.
Figure 2. Value of Multi-CoS in Carrier Ethernet (graphic provided courtesy of Tellabs).
This capability is important because the opposite of oversubscribed is overbuilt (to keep making the Ethernet service larger because of congestion times). Overbuilt means higher cost and less efficiency. With CE 2.0 certified Multi-CoS, the financial transactions are flawless … and don’t worry, the party invites will arrive in everyone’s inbox, since the corporate mail program just retransmits a few seconds later when more bandwidth is available.
Improvement 2: Enhanced OAM
There is one thing more frustrating than an interruption in your Ethernet service: An interruption in Ethernet service and not knowing who is responsible or why it was interrupted. Many Ethernet services that connect remote centers are supported by multiple providers and multiple vendors’ equipment. During an outage, delays in restoration may keep service down longer than anyone wants because of limited OAM tools. Too often this lack of information derails into high-stress vendor/service -- provider/customer meetings and conference calls. These occur when everyone is trying to determine which port, which fiber or cable, or which device was at fault, and how long it will take before it’s fixed.
The enhanced OAM, defined by the MEF, provides a lot of things, but of greatest value may be the ability to isolate faults in the network. CE 2.0 certification tests for enhanced OAM and quickly determines where problems are occurring and the nature of the problem. This capability is meant to curb the discussions of who is at fault for a service outage and expedite service restoration to improve customer satisfaction.
Improvement 3: Interoperability and E-Access
There was a time long ago when every service provider had its own set of wires into a building. While there are cases where this may still make sense, most end buildings have a small set of connections -- fiber, coax or twisted pairs. Some service providers offer their connections for leased access such that other service providers can gain access to the building without laying their own fiber, obtaining right of ways and spending the time to build this access connection. This leased-access approach has worked since the days of divestiture, with T1 being the dominant access technology.
As packet services and bandwidth expectations have increased, the days of time division multiple access (TDMA) are numbered. However, there have been impediments to Ethernet’s role in wholesale access, principally a lack of standardized external network to network interconnect (E-NNI) between service providers.
The E-NNI is the key component of CE 2.0 E-Access certification. (See Figure 3.) Now communication service providers may know exactly what to expect from wholesalers that are CE 2.0 certified -- a set of Ethernet features that assures end users’ SLAs will be met. This creates an environment where Ethernet services can be established very quickly, whether the service is local, regional, national or global. This interface definition will further enable the rapid growth of packet services across the globe.
E-Access may be similarly applied to the mobile backhaul and other applications where a different provider provides access.
Figure 3: E-Access example (graphic provided courtesy of MEF).
Assuming that you’re sold on the enhancements that CE 2.0 will bring, the follow-on question is why now? If what you have is working, there’s a lot of wisdom in leaving well enough alone. As the saying goes, “If it ain’t broke …”
Looking at the landscape of data communications services, it’s hard to see any area that doesn’t need more bandwidth, security, reliability and resiliency. Cloud services, both public and private, are appearing everywhere. Web-based services are staples to any business operation and they are steadily requiring greater resources with every push from the IT department. Remote data centers and satellite offices require more bandwidth and reliability. For those companies that compete in providing these services, any differentiation in Ethernet services has increasingly been limited to cost: “How many bits per penny can you give me?” The timing of CE 2.0 is ripe for early adopters to distinguish their services from competitors. And it’s healthy for the industry to compete on more than just price.
Ethernet is well on its way to becoming the power plug of the networking world (i.e., it plugs in and works the same way wherever you go). CE 2.0 is a push in the right direction for Carrier Ethernet and those that benefit from those services. Certification reduces the guesswork for those that need to deploy and support this equipment because the equipment and service provider have certified their compliance in advance. That leaves a lot less room for the double talk that sounds like, “Oh yes, we kind of do that … it works a little different … it’s a proprietary method … it’s just as good as …” These have to stop.
As for the next time you need to address any kind of RFI or RFP, etc., you may expect the following format: Question #1) MEF CE 2.0 -- Yes? No? If Yes, continue with RFP. If No, thanks for your time. That’s all we needed to know.
Welcome to the next age of connectivity.
Eric Geelen is a systems engineer for Tellabs and a Tellabs representative to the MEF. He has more than 15 years of experience in IT and telecom.
Tom Rarick is product manager for Tellabs optical product group. He has more than 25 years of experience in Ethernet and telecom.
For more information, please email firstname.lastname@example.org or visit www.tellabs.com.
What’s your take on this subject? Leave a comment and get the conversation going.