I had a great time on the Metro Ethernet Forum Q2 meeting in Munich last week. This was in part a result of the YANG models for PM and FM that I’ve been working on for a while passing letter ballot (they will be published any day now), but also because it seems that the direction of the forum is shifting towards something I feel very strongly about; manageability.
As you may know, the MEF organization and membership concerns themselves with defining and promoting the use of Ethernet for various types of interconnection services. They do that by not only planning, designing and writing technical specifications, but also by an interesting range of other activities to support the proliferation of these standards including certification, marketing and hosting ethernet service provider get-togethers.
The event was co-located with Light Readings Ethernet Europe event. Listening in on some of the keynotes from the the likes of Deutsche Telekom, AT&T, TeliaSonera and Equinix it was evident that there are pretty massive investments being made into replacing traditional leased-line interconnect services with Carrier Ethernet.
The defining moment for me was when one of the technical committee chairs looked at the agenda and only half-jokingly said:
We should rename ourselves the ‘Managed Ethernet Forum!
He was referring to the fact that with a few deviations, most of the cycles spent in the technical committee are now dedicated to various aspects of manageability of Ethernet services. In a way, this is a testament both to the quality of the technical foundation of these services and to the breadth of productifiable (!) features that they provide. All the service level features for providing a complete portfolio of connectivity products to the enterprise market are now in place. The next obvious step is go beyond early adopter-deployments and bring this into (very) large scale deployments. This is where the management aspects of the standards become critical.
The various current vendor implementations of the MEF standards provide a rather daunting challenge for operations managers looking to streamline their network. The number of knobs to twist and levers that are expected to be twiddled in the right order to provision, activate, test, dynamically change, and tear down services are many and involved. The lack of interoperable implementations of interfaces to do this across the wide range of vendors makes this a nightmare to manage at any ambitious scale without falling into the tar pit of proprietary scripts or vendor specific management systems.
The MEF leadership clearly hears operators complaints around this lack of scalable, robust and standardized service life cycle tooling. The MEF recently announced “Carrier Ethernet 2.0” and it is all about “multi-CoS, interconnect, and *manageability*”. I’m now eagerly looking forward to the more detailed discussions on the content of “Carrier Ethernet 3.0” as I know that keywords like automation and software defined are already being thrown around.
Get in contact if you want to discuss the manageability of carrier ethernet services both on the element side and on the element management side. Tail-f has a whole set of solutions aimed at helping network equipment providers rapidly build industry-leading configuration management systems based on open standards.