Michael Brenner is the senior vice president for product strategy at ClearPath Networks. He is responsible for product management and driving product direction. Michael is also a Vice Chair of the ETSI NFV ISG and recognized as one of the leading contributors to the ETSI NFV Architectural Framework and management and orchestration (MANO). In this interview with SDxCentral, Michael will share insights on one of the most important network functions virtualization (NFV) use cases virtualized CPE (vCPE), and the technologies enabling scale out and distributed NFV.
This second interview complements a previous interview with ClearPath Networks founder and CEO, Cliff Young.
SDxCentral: ClearPath was founded in 2002, and has been one of the pioneers of cloud-based networking. Recently, ClearPath has focused on virtualized CPE (vCPE), identified by carriers as one of the highest priority use cases for NFV.
Can you comment on the company’s history and how ClearPath evolved to address NFV and vCPE?
Brenner: Since the company’s inception, ClearPath has a long history of innovating in virtualized service delivery. ClearPath invented cloud-managed networking, and in 2008 it commenced the first product deployed leveraging OpenStack with a Tier 1 carrier in Europe. A few years later, we introduced our Virtual Services Platform (VSP) in 2011, more than a year before NFV was launched.
For the past few years, we has focused on the virtual network function (VNF) as a Service (VNFaaS) use cases defined by the ETSI NFV ISG, and on the vCPE use case in particular. An IHS/Infonetics survey of global operators revealed business vCPE as the #1 NFV use case, because of the potential to drive new revenues, while simultaneously reducing costs. Today we are capitalizing on our experience and evolving our technology to enable the lowest cost, personalized services for a range of scale out use cases.
In our recent report on the Virtualized Edge, dozens of firms responded with vCPE and SD-WAN offerings. How is ClearPath differentiating itself in an increasingly crowded field? What use cases is ClearPath pursuing?
Brenner: We are not at all surprised by the growing interest in vCPE and SD-WAN, considering the operator’s excitement about the potential to improve agility and revenue while simultaneously driving down both opex and capex. However, there are many distinct vCPE approaches, including firms focusing on VNFs, CPE providers enabling Distributed vCPE, and CO/POP-based solutions to name a few.
ClearPath has set its sights on enabling operators to move down-market to deliver personalized services at the lowest cost to their small and medium business (SMB) and consumer customer bases. Traditionally, operators have been less successful leveraging their managed services successes for SMBs/consumers, where full-featured services are priced out of the market.
Typically, vCPE approaches instantiate 1 VNF per VM (or multiple VMs), which are service chained over the network, and require a vendor-specific VNF Manager for each VNF. ClearPath is challenging the status quo to realize the vCPE value proposition through a next generation platform and a different way of packaging VNFs that addresses these shortcomings.
Considering the price sensitivity and the required scale, enabling operators to serve the SMB market is a tall order. What are the most important technologies ClearPath is developing to achieve that goal?
Brenner: ClearPath is pursuing a number of technologies that work in concert to enable personalized services delivery at the lowest cost. First, we adopted containers to minimize the NFV infrastructure costs, which is a significant contributor to overall service delivery costs at scale. We also optimized an extremely compact, embedded Linux OS that consumes a fraction of resources compared to a traditional distribution.
Our containers are capable of hosting a broad-range of VNFs. For the SMB/Consumer market, VNFs must have a reduced footprint and price point and open source VNFs offer a cost-effective alternative to large and complex commercial VNFs suitable for operator’s infrastructure and large enterprise.
And perhaps most importantly, we developed an open Virtual Services Platform (VSP) that includes VNF Manager (VNFM) functionality capable of LifeCycle Management integrated with traditional FCAPS management. Our VSP is designed to alter the ‘1 VNFM/VNF’ paradigm that has been a major obstacle to widespread vCPE adoption (and NFV in general), and change the notion that FCAPS management should be segregated from VNF Lifecycle Management.
It appears that what you term as NanoServices are critical to the ClearPath solution. Can you describe NanoServices in more detail including the benefits? Are they related to Microservices?
Brenner: ClearPath adopted the term NanoServices to describe our unique service building blocks, which are key to achieving personalization at scale.
NanoServices, are the fundamental service building blocks operators may use to deploy extremely cost-effective and personalized services.
Each NanoService consists of multiple VNFs, service-chained together at the Application layer (intra-container) that reside in a single container. NanoServices may be deployed on COTS CPE, or provider-side edge cloud, and may be service chained together over the network to enable operators to compose flexible services. VNFs may be dynamically injected or removed to/from a container during runtime to open the door to a new dimension of intelligent and dynamic services.
Microservices have been traditionally used to describe application software architectures based on modular building blocks that communicate through distinct APIs, but there is no standard definition. While similar, we were seeking a more precise, and less ambiguous term to capture all aspects of our service building blocks, and draw attention to the minimized footprint.
You mentioned that ClearPath is concentrating on scale out use cases. What do you mean by that? Can you provide an example?
Brenner: Examining the overall NFV space, we propose a use case classification based on the types the VNFs being delivered. Way back in 2012-13, when NFV was taking shape, NFV pioneers set their initial sights on virtualizing software embedded in thousands of purpose-built appliances that operators used to run their network.
Virtualized versions of the hardware-based appliances were full-featured, relatively large, and centralized. These virtual appliances were typically deployed in one (or a few) VMs that served thousands and even millions of end users. Subsequent generations of VNFs are becoming more efficient as the industry gains experience with NFV. Because of the need, for the VMs to adapt to elastic demands, such use cases are referred to as scale up.
Scale out use cases address a different problem. Instead of deploying large and complex VNFs in a single (or few) VMs, scale out use cases are decentralized; service chains consisting of relatively small, and granular-functionality VNFs are personalized and deployed using virtualization containers to support groups, or ultimately individual subscribers. Services are dynamic, and tailored to the needs of operators’ individual customers.
An example of scale out use cases is described below for mobile operators seeking to enhance value, reduce churn, and maintain profitability. Instead of offering basic mobile data services, an operator can offer a secure mobile data service that can be tailored to the needs of various customers, ranging from consumers, power-users, SMBs, etc. Each group or individual customer may obtain the specific services it requires (and willing to pay for), without affecting services delivery to other customers. Security capabilities can be offered through open source VNFs and combined with other VNFs to improve security, performance, monitoring, troubleshooting, etc. As a result, operators are able to raise the bar versus the competition, rationalize higher pricing, and at a minimum, stem price erosion and improve stickiness.
ClearPath has been involved in a number of high-visibility proofs of concepts (PoCs) this year and has appeared on partner lists of prominent infrastructure vendors. What can you share with us about them?
Brenner: Indeed, ClearPath has been engaged with the world’s largest and important operators in a series of PoCs showcased at major events:
- AT&T /Ericsson/ClearPath PoC OPNFV Summit – In this PoC, ClearPath teamed up with AT&T and Ericsson to demonstrate a software-defined WAN (SD-WAN) enabled by software-defined networking software-defined networking (SDN) that enabled legacy to SDN interworking.
- Orange/Intel/ClearPath OPNFV Summit – In this PoC, ClearPath teamed up with Orange and Intel to demonstrate secure, managed services for SMBs.
- AT&T/China Mobile/Dell/ClearPath PoC NFV World Congress – In this PoC, ClearPath teamed up with AT&T, China Mobile, and Dell to demonstrate VNF migration across federated OPNFV
In all three PoC demos, ClearPath’s VSP platform managed multiple vCPE instances, providing integrated VNF lifecycle management with FCAPS management support to improve overall agility, while driving down both capex and opex.
What can we expect from ClearPath in 2016?
Brenner: 2015 has been quite a year for ClearPath, as we shored up our leadership team, forged partnerships with industry leaders, engaged with major cable, carrier, and mobile operators, and made considerable progress on our flagship VSP platform initially optimized for vCPE.
Looking ahead, ClearPath is planning a major VSP release in Q1-2016, along with maintaining our growing customer engagements (most of which we are not at liberty to discuss). Another important initiative is our Model-Driven Automation project. Our goal is to address the deployment hurdles for vCPE/VNFaaS- the cost/complexity/time required to onboard deploy, and manage new VNFs, along with integration of our VSP platform with the proliferation of operators’ back-end OSS/BSS/Orchestration/Management systems.
In 2015, ClearPath contributed an open source project YangForge, a YANG-based tool designed to model VNF behavior to facilitate on boarding and management. YangForge was adopted by the OPNFV Promise Project for resource reservation and management, predicated upon an information model for NFV resources. Look for a product introduction of YangForge in 1H2016.