Colt Technology Services is working toward launching intercarrier services early next year with multiple service providers.
During an interview at the MEF18 conference in Los Angeles, Colt's Mirko Voltolini, head of Network On Demand, said the original target date for the new Carrier Ethernet services with other service providers was next month, but now that has been pushed back a bit.
"We'll always remain optimistic about time schedules, but I think between now and end of Q1, we're probably going to be able to launch with at least three if not four carriers," he said. "It's all coming together at the same time."
FierceTelecom previously reported that Colt and Verizon planned to launch two-way, intercarrier network orchestration for services across each other's networks. Using a programmatic API-to-API interface between their separate SDN architectures, Colt and Verizon will use automation to provision a Carrier Ethernet service across their production networks.
Colt and Verizon are using MEF's Lifecycle Service Orchestration (LSO) Sonata APIs for intercarrier services. MEF announced a developer release at its trade show last month.
While the service with Verizon isn't officially in production just yet, Voltolini said both service providers have the LSO Sonata API working in production. During MEF18, AT&T was mentioned as one of the carriers that Colt would be working with on intercarrier orchestration for Carrier Ethernet services next year.
As for a third carrier, Orange Business Services would seem to be a likely candidate based on the fact that it has worked with Colt, Orange and AT&T on various proof of concept demos.
At MEF 17 last year, AT&T, Orange, Colt, and a handful of vendors showed a proof of concept demonstration using MEF's LSO Sonata software development kit (SDK) to provision Ethernet services across AT&T and Orange's SDN architectures.
At last year's Mobile World Congress in Barcelona, AT&T, Orange Business Services and Colt announced they had formed an orchestration partnership that would allow them to provide Carrier Ethernet services across each other's SDN architectures using TM Forum APIs and MEF LSO APIs.
Enabling cross-carrier automation of a wholesale service between two service providers is a game-changer for the industry. Epsilon and DCConnect announced in June that they would commercially deploy their two-way, inter-carrier network orchestrations service in the third quarter of this year.
In this Q&A, which was edited for clarity and length, Voltolini talks about the importance of LSO Sonata, as well as other MEF LSO APIs, what it has been like serving on MEF's board of directors and the importance of MEF releasing the first draft of its SD-WAN service definition and certification.
FierceTelecom: Earlier this year, you were named to MEF's board of directors. What's your impression so far?
Mirko Voltolini: We've been working with the MEF now for a number of years, but I was never deeply involved personally. I used to send people from my team and we've been actually working with MEF since Carrier Ethernet 1.0, so it has been 10 years or so.
MEF really has transformed into a much more larger organization than I remember in the past. What prompted me to apply for a board position was that I wanted to get involved and try and change the direction for the industry.
The work we did with Sonata, which started with us and AT&T primarily in the beginning and then Orange joined swiftly, was the key point where we thought that we really needed to get the industry to try and change how we work with each other. I can't really tell you much about how it feels to be on the board because it's only been a few months now. I am pleased that I can be involved in key-decision making. But in particular, I want to help drive this agenda around the adoption of Sonata, which for us is a fundamental thing. We sell and buy with other carriers and it could really change the shape of our propositions if we are able to make the delivery across third-party networks.
So that is kind of the main motivations behind me joining. Obviously I support the expansion MEF is doing into other service domains as well. We should not try to conquer the world all at once, but there are lessons learned that at least could be generalized and extended to other services.
FierceTelecom. Before you joined the MEF board, you were a real vocal proponent of LSO Sonata. Where are we now?
Mirko Voltolini: So we have a draft release and I think it's great. I can also say we are pleased with the fact that we talked about Sonata a year ago MEF 17, and a relatively short amount of time, one year, we now have a draft release out that can be used by carriers. It's not yet fully approved, but it is already in a form that can be used by carriers. So I'm pleased by the fact that it's not taking forever to develop these capabilities.
FierceTelecom: For now, LSO Sonata has four elements—serviceability, product inventory, quoting, and ordering—but there will be more?
Mirko Voltolini: It will also need the fulfillment billing capabilities and capabilities for trouble ticketing. But I think it's good enough to form the shape of the initial API that the carriers can adopt it. There are certain things that maybe we will continue to do on top of it. Today, with another billing interface, we can use our own appropriate API to cover 80% of the lifecycle of service billing and we can continue to use the Colt API until Sonata is ready.
FierceTelecom: So aside of Carrier Ethernet services, MEF said it was also working towards LSO Sonata-enabled carrier to carrier SD-WAN, IP services and optical services. What is the time frame for those?
Mirko Voltolini: SD-WAN is important because we don't have the set standard yet. Everybody has got his own SD-WAN solution right now. Integration with third parties will become important in the near future, I would say probably in a couple of years' time. We could benefit from that end-to-end integration.
But SD-WAN has got more problems than Ethernet; it's not just intercarrier interconnection. It's also that the services have been built in different manners by different vendors. So there is a protocol level and harmonization that need to happen first. And it's not going to happen very easily because everybody has got his own solution. You cannot go and tell Versa, Silver Peak and VeloCloud that they have to implement the same standards, because right now the standard doesn't exist. So, first thing is to define the standard and common terminology, which is what MEF has done.
I would say if we have a Sonata-like integration between SD-WAN and Windows in a couple of years time, and we have a common definition of the SD-WAN protocol level, then we'll really be in a good place.
FierceTelecom: What about other LSO APIs such as Presto, Interlude and Cantata?
Mirko Voltolini: Cantata is like Sonata. Cantata is between a service provider and an enterprise. But it's like Sonata with a couple of minor variations. If it were up to me, we probably would have called it Sonata and avoided having two different interfaces. Cantata is also important because large enterprises, and you see them on the executive enterprise council that MEF has set up, also want to use APIs to control services. They don't want to use all of the forms anymore. They have their own systems and so common terminology will definitely be useful for that interface. So that's useful and important for us.
Presto and Interlude are a little less important for us. Presto is downward, so north to south. It would be nice if it could have all the vendors implementing the same interfaces but it's not like we're dying for that. Today I would there are too many interfaces. You have someone implementing RESTful APIs and then somebody else implementing another standard. There are multiple options. But we can work with that for the time being. Presto is north/south, so that is a question between the vendors and the OSS systems.
Interlude is also less important. There are some use-cases where you should not use Sonata, because Sonata is business-to-business. Interlude is controller-to-controller. We can do a couple of things without Interlude; we don't have it. We can still use Sonata even if it's not a business case. We can use Sonata to communicate, or we can implement a substitute of the Interlude API if we really need to talk directly to the third-party controller.