1:30 PM -- I recently had the chance to chat with Zeev Draer, vice president of strategic marketing for MRV's Optical Communications Systems division. Draer has more than 20 years of experience in the industry and participated on mobile backhaul and network edge panels at the recent Ethernet Expo Americas.
Below are highlights of my interview with Draer:
Stan Hubbard, Heavy Reading: What do you see as the major big-picture trends shaping the industry right now?
Draer: We are at a maturing stage in the networking industry, with the market evolving in an exciting direction that aggressively transforms from legacy Sonet/SDH technologies and changes service providers' network strategies. It's not about big pipe connectivity, but more about application-driven intelligence that must be empowered with strong end-to-end multi-layer provisioning of services, performance monitoring across layers and high elasticity of the network that should scale to millions of subscribers and services.
In terms of infrastructure, there's no doubt that fiber rollout in access networks has a strategic impact on the spectrum of services that can be introduced, and wholesale mobile backhaul is fueling more fiber-related investment. In parallel, mobility and cloud applications are playing a phenomenal role in design and implementation strategies that drive new technologies with disruptive impact. We are seeing more trends like packet inspection, caching and dynamic bandwidth allocation that empower the network to use intelligent decisions for traffic management, to increase the number of service options and to create differentiation that can be monetized.
Heavy Reading: What are your customers asking for most when they plan to upgrade their metro networks to support the business, residential, wholesale and/or mobile backhaul services?
Draer: Our Tier 1 customers want to be able to scale their networks and utilize more services over the same infrastructure with end-to-end visibility in various layers as well as automated provisioning. There are commonalities and there are differences in terms of networking features, but the fact is that service providers can reuse the same platforms for business, mobile backhaul and triple play aggregation without compromising on performance. They also can use exactly the same operating software and common application-aware hardware with powerful service provisioning that can handle millions of devices and services.
Heavy Reading: Our research shows that two of the highest priorities for operators are to accelerate the rollout of their services and to more effectively manage those services by leveraging service OAM. What are you doing specifically to address these two concerns?
Draer: I certainly will concur with Heavy Reading's research findings. Service providers are exploring strategies to address their need to create value in light of price erosion in a mature, highly competitive telecom market.
Our strategy for service provisioning and management is a reflection of a five-year plan dating back to 2008 that shaped very tight cooperation with Tier-1 customers and evolved into a very robust product that is named Pro-Vision. It's fully compliant with MEF Carrier Ethernet 2.0 specifications, includes an up-to-date service OAM toolkit and cloud-based customer portal, and is optimized for large-scale operations. Pro-Vision enables a very efficient, centralized and automated turn-up process, service provisioning and SLA monitoring. These building blocks are the cornerstone of our innovations that truly are helping operators differentiate themselves versus their competition.
Heavy Reading: Do you have any data points that you can share on how much operational cost savings your customers have been able to achieve by utilizing OAM capabilities in their networks?
Draer: It's a statement of fact that new networks are SLA-oriented. Performance-guaranteed services are priced higher and customers are willing to pay more for them, especially in cloud access networks that have business SLA dependencies. But service providers must have a rich set of OAM capabilities to measure network KPIs and detect performance degradation before it reaches the point of violating an SLA and involves penalties as well as an overall increase in opex.
One of the ways for service providers to compete is to reduce opex and increase service lifecycle efficiency. We offer to our customers a complete OAM framework that translates into a number of proven benefits: an automated service turn-up process with a zero-touch concept; reduced technician on-site operations and time to repair; fewer tests required in the field; end-to-end service visibility; and ultimately, higher reliability and better QoS with performance statistics that can be shared with Web portals. The net result is reduced opex of more than 50% versus manual configurations, site visits, and SLA penalties and an overall increase in operational efficiency. In addition, we offer proven interoperability to OSS back-office systems that's critical for automating the Ethernet service lifecycle, reducing expenses, speeding up the introduction of new services and helping service providers increase revenue.
Heavy Reading: I'd like to explore this idea of multi-layer management. What multi-layer capabilities do you offer today and how do you see that evolving?
Draer: With the Pro-Vision platform, our portfolio has the ingredients to serve IP, MPLS, Ethernet and optical layers and orchestrate all the complexity. Our OptiSwitch and Master-OS Linux control plane serves as a multi-layer application-aware transmission platform that, when accompanied with our LambdaDriver and FiberDriver optical transport and Pro-Vision platforms, can scale to support millions of devices and emerging services and applications. As a whole, this solution enables service providers to function beyond dumb pipes and create true value. Each network layer is "stamped" differently and requires very tight design planning, OAM at different layers with measurable KPIs and correlation alarms to match service integrity. Binding our paces with Carrier Ethernet 2.0, we use the latest MEF service OAM standard definitions, Ethernet ring protection based on the ITU-T standard, and add-ons for simplified automated topology discovery, end-to-end service provisioning and SLA reporting via a Web portal.
Over the next few years we will be able to add more intelligence from higher layers and create more application-awareness and service elasticity in our entire portfolio that will automate dynamic service ordering as well as support the cloud access services with better quality of experience that will continue to aggressively proliferate.
Heavy Reading: Do you care to comment on the SDN trend? Does MRV have any SDN-related plans that you can share?
Draer: I strongly believe that the impact of Google adopting the SDN concept created a massive ecosystem around the entire industry. SDN is not really a new concept. Central control planes were around in the Sonet/SDH days, so it's not a complete revolution. But SDN is an evolution that commoditizes equipment and transforms the industry to a software and applications orientation. The question is not if SDN will take off, but how, where and when. In "wall-gardened" networks like CDNs this might be more practical today, and why it was implemented pragmatically by Google. However, it's more complicated for service providers -- especially Tier 1s -- that have very tight OSS processes and will not embrace new concepts that disrupt operations. The complexity of multiple layers requires the networks to keep existing protocols of distributed IP control plane and it's more complex than just to say: "Lets implement the OpenFlow protocol on networking devices and place an SDN controller in the cloud." Nonetheless, there's good progress in SDN and the latest OpenFlow v1.3 incorporates an IPv6 scheme, so it's getting there.
MRV's strategy is two-fold: adding a Path Computation Element (PCE) tool to enhance centralized network control in the Pro-Vision software and to set interoperability and linkage to the emerging OpenFlow protocol in future stages. PCE is part of the SDN framework and a well-established client/server IETF standard that complements the SDN story and solves some real-world problems, so we feel confident in executing our direction in stages, but we'll follow carefully the market evolution and customer requirements.
Heavy Reading: You talked about access to cloud services and the challenge of secure access. Can you share your thoughts on this matter, and does MRV have any security-related plans that you can share?
Draer: The emergence of cloud-based services through remote access introduced and created a real challenge and complexity in secure authentication and identity management. Authentication is required in order to verify the integrity of a user and specifically to linkage this access to networking resources.
Most of the security solutions provided by technologies currently available in the market are expensive due to constant customer support needs, such as: installation problem issues, changing technologies, bad user interface experience and divergence of operations.
We are in process of exposing our new CloudScreen solution that's built on top of the company's award-winning OptiSwitch intelligent Ethernet demarcation service series. This new concept streamlines multi-layer authentication technology and provides enhanced security authentication and identity management and consequently prevents identity thefts both from the service provider and from the user and can control access exactly at the demarcation point of the network before security breaches penetrate service providers' business services.
-- Stan "EtherMan" Hubbard, Senior Analyst, Heavy Reading