Accelerating the Transition of Open Source MANO from Proof-of-Concept to Production

The history of Linux is that it’s a technology that started as a humble hobby that now enjoys more than 75 percent marketshare among enterprises as the primary cloud platform. At least according to one source, the driving force behind this open-source juggernaut was Red Hat building a highly profitable business around its Linux distribution, an industry first. By doing so, Red Hat not only proved the market viability of Linux, it significantly raised its overall profile of open source.

RIFT.io’s ambition is to achieve that level of impact eventually with our active participation in the OSM project whose stated mission is to deliver production-quality open source MANO stack that meets the requirements of commercial and carrier-grade NFV networks. The project has already delivered on two significant milestones by making Release 0 and ONE codes available all within six months since the inaugural meeting of the OSM community. This week, RIFT.io took a major step toward demonstrating market viability by delivering the industry’s first commercial software distribution of OSM in our latest version of the RIFT.ware platform.

The customer value proposition behind this new RIFT.ware release is in large part to accelerate market adoption of OSM and NFV through advanced, carrier-grade features that simplifies the onboarding and management of VNFs wherever they’re deployed. RIFT.ware achieves this as an open, simple, automated, and cloud-scale platform to simplify the management and orchestration of virtual network services using the agility of open-source innovation including OSM, of course. Looking at this graphically, we can argue that this is a major step toward moving the entire NFV industry to the next phase of its maturity as we like to show using this chart:

As the first commercial software distribution of OSM, RIFT.ware 4.3.3 delivers significant carrier-grade features and functions that will accelerate progress along the NFV maturity model toward a world of open, hyperscale, cloud network services.

 

It’s a significant step to be sure but we have a long ways to go, even according to our own optimistic estimations. I got a first-hand glimpse into the challenges ahead at the recently held SDN World Congress where the OSM held a technical steering committee/end-user advisory group OSM workshop, which attracted almost 100 enthusiastic attendees from leading operators and vendors interested in NFV. I wrote my observations from the sessions for the benefit of the OSM community in this blog but here’s a further distillation of the major challenges:

  • Getting mobile operators and SPs in general to think beyond just industry standards and getting them comfortable with the new and exciting possibilities of open source remains a critical requirement. This recent Light Reading column did a nice job of summarizing the pros and cons of both standards-based and open source development, even going as far as making the case that the trend is tilting toward open source. This is why the OSM being an operator-led project is more than just nice-sounding Marketing; it’s important to keep them interested and, more importantly, actively engaged,
  • Simplifying VNF deployment and management is a must-have feature and functionality. As a new network paradigm, NFV has a high-enough hurdle in earning the trust of operators and SPs. Any process that is overly complex is likely to elongate the market adoption timeline that much further. Thus, RIFT.io believes that it’s important for vendors to shift the burden of solving complex MANO problems from the operators and VNF builders back to them. One tangible way we accomplished this was with the VNF Descriptor Package Generator available for both RIFT.ware and OSM users in order to speed up the VNF deployment process without having to modify the VNF itself,
  • Enabling operators and SPs to attain “cloud speed” on their own There’s no question that while NFV is itself leveraging IT concepts and technologies to make operators and SPs more agile and responsive, success will require these technologies to be deployed in an SP-centric, not IT or even cloud-centric, way. Therefore, the value for platforms such as RIFT.ware will be in their ability to bridge the worlds of telecommunications and cloud-native architecture and apps seamlessly. In other words, while the services may be IT-based, they have to “talk telco”. The good news for RIFT.io customers is that this what we as a company thrives in doing, as our friends at Canonical graciously pointed out at SDN World Congress

In summary, RIFT.ware 4.3.3 as the first commercial distribution of OSM is about bringing our proven expertise in onboarding, service composition, and network service delivery toward building a production quality, commercial software. Our experience in this area will pave the way for even more ambitious challenges and solutions in the next releases of RIFT.ware and OSM.