OPNFV Book Preview #3: All Roads Lead to OPNFV

By | Blog

This post first appeared on

By Amar Kapadia

Previously in our discussion of the Understanding OPNFV book, we provided an introduction to network functions virtualization (NFV) and explored the role of OPNFV in network transformation. We continue our series with a look at chapters 4 and 5, which provide a comprehensive description of the various open source NFV projects integrated by OPNFV and the carrier grade features contributed back to these upstream projects by the community. In this article, we cover these two topics briefly and provide some related excerpts from the Understanding OPNFV book.

For those less familiar with OPNFV, according to the OPNFV website:

Open Platform for NFV (OPNFV) facilitates the development and evolution of NFV components across various open source ecosystems. Through system level integration, deployment and testing, OPNFV creates a reference NFV platform to accelerate the transformation of enterprise and service provider networks.

OPNFV is the only open source project that integrates, deploys, and tests a wide range of open source NFV projects on a continuous basis. If you believe that open source is the future of NFV, then OPNFV is a project you definitely want familiarize yourself with.

The list of projects integrated by OPNFV includes the following categories:

  • NFV infrastructure (NFVI)
    • Hardware
    • Virtual compute
    • Virtual storage
    • Virtual networking and dataplane acceleration
  • Virtualized infrastructure manager (VIM)
  • SDN Controller
  • Management and network orchestration (MANO)

More details on the various projects in each category are outlined in the book. For example, here is an excerpt on the OpenDaylight project:


Like OPNFV, OpenDaylight (ODL), is also a Linux Foundation project. It is a full blown modular SDN controller that caters to multiple use cases such as NFV, IoT, and enterprise applications. It supports numerous southbound interfaces to manage virtual and physical switches (OpenFlow, Netconf and other protocols). For the northbound interface to OpenStack or other orchestration layers, ODL uses YANG (a standard modeling language) models to describe the network, various functions, and the final state. The ODL community is large, with Brocade, Cisco, Ericsson, HPE, Intel, and Red Hat being just a few of the companies supporting the initiative.

In addition to integrating upstream projects, the OPNFV community plays a critical role by identifying carrier grade feature gaps, developing code to fill those gaps and contributing the code back to respective upstream projects. The book discusses 24 OPNFV feature development projects and groups them into the following four categories:

  • Service assurance and availability
  • Easing integration of upstream projects
  • Deployment and lifecycle management
  • Documentation and security

The book describes each of the 24 projects. For  example:

One of the more important feature development projects is OPNFV Doctor that provides an  NFV service assurance framework. As with other feature development projects, the OPNFV Doctor project develops and contributes service assurance features directly to the upstream projects, in this case OpenStack Congress, Nova, Neutron, and Cinder.

Want to learn more? You can check out the previous blog post that discussed the broader NFV transformation complexities  and how OPNFV solves an important piece of the puzzle, download the Understanding OPNFV ebook in PDF (in English or Chinese), or order a printed version on Amazon.

OPNFV Co-Hosts The Linux Foundation’s Inaugural Open Source Networking Days

By | Blog

This post first appeared on

By: Arpit Joshipura

One of my primary goals at The Linux Foundation is to foster innovation across the entire open source networking ecosystem. This involves coordinating across multiple open source projects and initiatives and identifying key areas for collaboration to create an open source networking stack.

We are working across the entire ecosystem with industry-leading partners — from developers to service providers to vendors — to unify various open source components and create solutions that will accelerate network transformation. As part of this journey, I am pleased to introduce Open Source Networking Days (OSN Days), a series of free events that are hosted and organized by local user groups and The Linux Foundation members, with support from our projects, including DPDK,, ONAP, OpenDaylight, OPNFV, PNDA, and others.

OSN Days are a fantastic opportunity for network developers and users to learn how ONAP, OPNFV, OpenDaylight  and other open source initiatives are changing NFV/SDN orchestration and networking solutions. Cities on the tour include: Paris, Milan, Stockholm, London, Tel Aviv, and Japan. Register today for an upcoming OSN Day in your region.

The day-long events will start with a plenary session where attendees will hear from site hosts and The Linux Foundation speakers on the state of the industry and the collaboration and touch points between projects that make up the open source networking stack. Presenters will also explore how business opportunities like 5G and IoT are enabled by network transformation.  In the afternoon, events may feature technical sessions, tutorials, demonstrations, and workshops that empower attendees to participate, contribute, and deepen their knowledge of open source networking.

Our first OSN Day kicks off October 9 in Paris, followed by stops in Milan (October 12), Stockholm (October 13), London (October 16), Tel Aviv (October 19), and Japan (October 19). Thanks to our incredible site hosts and sponsors: Amdocs, ATOS, Cloudify, Ericsson, Huawei, Intel, NEC, Orange, Red Hat, SUSE, and Vodafone, along with our high-caliber roster of speakers, for helping to make these OSN Days a reality!

More details about the events, including site-specific agendas, registration info, and details on hotel and travel, can be found here: If you have any questions, or would like to host an event yourself in the future, please email

OPNFV Intern Spotlight: Aakash Kt

By | Blog

We have a thriving intern program and are pleased to welcome even more talented students to the OPNFV community! OPNFV interns work directly with the community and receive hands-on development experience in NFV. Each intern works closely with an active OPNFV developer as their mentor on a project that suited interest and community need. This blog series aims to showcase these interns and the projects they work on, the mentors who are helping with their professional development, and their experience working in an open source community to help  accelerate NFV.

If you are interested in becoming an OPNFV intern, please email .

About Aakash (in his own words):
I am an undergraduate student at IIIT-H (International Institute of Information Technology, Hyderabad, India), pursuing a major in Computer Science. My research interest mostly lies in Computer Vision, and I am an undergraduate researcher at the CVIT lab at my university.

I am very much interested in NFV, especially OPNFV, since it is disrupting the networking world.

How did you hear about OPNFV and what got you interested in this internship?
I first heard about OPNFV from one of my university seniors. When I looked into this in more detail, I found it quite interesting. The idea of NFV is something unique, and something that has a lot of benefits. Wanting to be a part of this revolution, and learning all about it, was the thing that interested me most – in addition to the fact that I will get to work with such an amazing community!

Can you talk about your experience working on an open source project? Any previous experiences you can share or key learnings from working on OPNFV so far?
This is my first experience working in open source! I am grateful that my first dabble in open source is here with OPNFV. The vibe of the community and the interactions, and everyday opportunities for learning is hard to ignore. My experience here has been superb!

What’s the best thing you’ve learned from your internship?
Apart from all the technical knowledge gained, I think the most important thing that I learned so far is how to work with a global team. My communication skills and ability to deliver my thought process and intentions clearly have vastly improved.

Who is your mentor and what’s the experience been like?
My mentor was Mr. Narinder Gupta and my experience with him has been phenomenal! His ability to look at a problem and analyse it in the right way is something that I admire. Interacting with him has taught me a lot, both in terms of technical knowledge and in terms of social skills.

What’s your advice to other aspiring open source folks out there?
Things may seem overwhelming at first, and it may seem like there are a lot of people talking about a lot of stuff which you have no clue about, but go on and be bold! Don’t be scared; ask questions if you do not know something. This is a  crucial step. Things will eventually fall into place, and then you, too, will talk about things you thought you never understood!

What gets you jazzed to work with open source? (e.g., listening to music, drinking coffee, chatting in IRC, etc.)?
I’d say listening to music.

Being in open source allows you to work from anywhere in the world and interact with lots of different people. What have you found most surprising about the open source developer community?
The helpful nature of everyone around, and the extent to which they offer help, is something I found extremely surprising. In a good way, of course!

What do you want to do next? What is your dream job?
Well, what I want to do next with respect to OPNFV is to keep contributing! On the dream job front, I have not yet decided that far ahead, but I am inclined towards research.

Book Preview, part 2: OPNFV’s Role in Network Transformation

By | Blog

This post first appeared on

By Amar Kapadia

The Understanding OPNFV book takes an in-depth look at network functions virtualization (NFV) and provides a comprehensive overview of The Linux Foundation’s OPNFV project. In this article, we provide some excerpts from the book and discuss some organizational elements required to make your NFV transformation successful. These best practices stress how both technical and non-technical elements are required, with non-technical often being more critical.

According to the project’s website:

Open Platform for NFV (OPNFV) facilitates the development and evolution of NFV components across various open source ecosystems. Through system level integration, deployment and testing, OPNFV creates a reference NFV platform to accelerate the transformation of enterprise and service provider networks.

As the demand for more robust and agile networks increases, organizations are looking to technologies such as NFV to bring the scale and flexibility of cloud architecture to networking. OPNFV is the only platform that builds and integrates elements — including functional support for Management and Orchestration (MANO) — across multiple end-to-end open networking stacks that meet the needs of end users to enable accelerated NFV.

What is OPNFV?

OPNFV is not a traditional software development project. Instead, OPNFV integrates a number of upstream projects and tests them. Of course, there is a software development component as well, but those contributions are to fill carrier-grade feature gaps in upstream projects. The project is organized around the following three pillars:

  • Integration: OPNFV integrates a variety of open source projects to address specific NFV requirements.
  • Testing: OPNFV tests the entire stack across a variety of NFV-specific parameters.
  • New features: For each upstream open source project, OPNFV serves as the vehicle for NFV requirements. By actively working upstream and providing a single voice for NFV requirements, OPNFV steers these open source projects to better serve the needs of NFV.

OPNFV is an integration project whose work involves testing, CI/CD, documentation, and more. If you are looking to deploy NFV across telco or enterprise networks using open source projects, consideration of OPNFV isa wise first step that could save significant time and effort when setting up and running integrated testing of open source components.

How to Benefit from NFV

When starting your NFV transformation journey, there are both technology changes and organizational changes to consider. Here are three new technology changes that organizations need to embrace to get the full benefit of NFV:

  • Model-driven architectures
  • DevOps
  • Cloud native virtual network functions (VNF)

Model-driven architectures convert manual infrastructure definitions and processes to code, allowing full automation of network service and VNF deployment, and subsequent day-2 monitoring and lifecycle management.

DevOps introduces continuous integration (CI), continuous testing, continuous deployment (CD), and continuous monitoring of VNFs.

Additionally, cloud native VNFs allow for faster development and the use of cloud architectures where reliable, scalable services can be built on powerful and economic, but sometimes unreliable industry standard hardware.

In this book, we also outline what organizational changes need to be put in place to effect the above three technology changes.

In planning your NFV transformation, you should consider the following best practices, based on the experience of other telecom operators. The concepts are explained in detail in the book.

  1. Clearly articulate goals.
  2. Build skills organically
  3. Agile instead of big bang
  4. Pace your journey
  5. Find use case
  6. Find executive sponsor
  7. Use dedicated teams
  8. Spread the knowledge

Why Consider OPNFV?

As you can see, NFV transformation is about process, organization structure, and skill set acquisition as much or more than it is about the technology. And, OPNFV is an unparalleled community with deep exposure to these aspects in addition to technology. OPNFV helps not only with technology, but also more broadly with the non-technical aspects of NFV transformation.

Intrigued? You can learn more in the previous blog that introduced NFV concepts, download the Understanding OPNFV ebook in PDF (in English or Chinese), or order a printed version on Amazon.

Book Preview: Understanding OPNFV

By | Blog

This post originally appeared on

By Amar Kapadia

If telecom operators or enterprises were to build their networks from scratch today, they would likely build them as software-defined resources, similar to Google or Facebook’s infrastructure. That’s the premise of Network Functions Virtualization (NFV).

NFV is a once in a generation disruption that will completely transform how networks are built and operated. And, OPNFV is a leading open source NFV project that aims to accelerate the adoption of this technology.

Are you a telecom operator or connected enterprise employee wondering which open source projects might help you with your NFV transformation initiatives? Or a technology vendor attempting to position your products and services in the new NFV world? Or perhaps an engineer, network operator or business leader wanting to progress your career using open source projects (case in point, in 2013 Rackspace stated that network engineers with OpenStack skills made, on average, 13 percent more salary than their counterparts)?  If any of this applies to you, the Understanding OPNFV book is a perfect resource for you.

OPNFV BookIn 11 easy-to-read chapters and over 144 pages, this book (written by Nick Chase from Mirantis and me) covers an entire range of topics from an overview of NFV, NFV transformation, all aspects of the OPNFV project, to VNF onboarding. After reading this book, you will have an excellent high-level understanding of what OPNFV is and how it can help you or your organization. This book is not specifically meant for developers, though it may be useful for background information. If you are a developer looking to get involved in a specific OPNFV project as a contributor, then is still the best resource for you.

In this blog series, we will give you a flavor of portions of the book — in terms of what’s there and what you might learn.

Let’s start with the first chapter. Chapter 1, no surprise, provides an introduction to NFV. It gives a super-brief overview of NFV in terms of business drivers (the need for differentiated services, cost pressures and need for agility), what NFV is and what benefits you can expect from NFV.

Briefly, NFV enables complex network functions to be performed on compute nodes in data centers. A network function performed on a compute node is called a Virtualized Network Function (VNF). So that VNFs can behave as a network, NFV also adds the mechanisms to determine how they can be chained together to provide control over traffic within a network.

Although most people think of it in terms of telecommunications, NFV encompasses a broad set of use cases, from Role Based Access Control (RBAC) based on application or traffic type, to Content Delivery Networks (CDN) that manage content at the edges of the network (where it is often needed), to the more obvious telecom-related use cases such as Evolved Packet Core (EPC) and IP Multimedia System (IMS).

Additionally, some of the main benefits include increased revenue, improved customer experience, reduced operational expenditure (OPEX), reduced capital expenditures (CAPEX) and freed-up resources for new projects. This section also provides results of a concrete NFV total-cost-of-ownership (TCO) analysis. Treatment of these topics is brief since we assume you will have some NFV background; however, if you are new to NFV, not to worry — the introductory material is adequate to understand the rest of the book.

The chapter concludes with a summary of NFV requirements — security, performance, interoperability, ease-of-operations and some specific requirements such as service assurance and service function chaining. No NFV architecture or technology can be truly successful without meeting these requirements.

After reading this chapter, you will have a good overview of why NFV is important, what NFV is, and what is technically required to make NFV successful. We will look at following chapters in upcoming blog posts.

This book has proven to be our most popular giveaway at industry events and a Chinese version is now under development! But you can download the eBook in PDF right now, or order a printed version on Amazon.

Communications Service Providers increasingly using OPNFV for NFV Deployments

By | Blog

By Sandra Rivera, Vice President and General Manager, Network Platforms Group at Intel Corporation, and OPNFV Marketing Committee Chair

During the recent OPNFV Summit in Beijing, we announced the results of our third Heavy Reading survey, aimed at gauging value of OPNFV as well as the state of NFV deployments among global communications service providers (CSPs). The highlights of the findings are provided in slides, which were presented by Heavy Reading analyst Roz Roseboro in a session during the event, but we would like to provide some additional insight into what we’re seeing this year as the industry evolves.  

Digging deeper into the survey data  was telling: While the importance of OPNFV to CSPs overall has increased over the past year, it’s become even more important for those organizations with NFV already in production70% of those surveyed, in fact. And most of those with NFV in production (86%), are actively following OPNFV, with many (64%) contributing to the project. Note: more than half of all contributors work at companies with NFV in production. Importantly, nearly a third of CSPs with NFV in production have plans to adopt an OPNFV stack as part of their NFV architecture. This tells us is that the work we’re doing is valuable and is being used in commercial solutions across the globe.

Evidence of this is also seen in a shifting interest in NFV technologies (particularly open sourced) that goes beyond software and into hardware. The Open Compute Project (OCP), which aims to apply the benefits of open source to hardware, ranked as the most  important upstream project to the OPNFV community (in addition to OpenStack and SDN controllers). This signifies a turning point; it’s an indication that NFV is getting closer to deployment as CSPs acknowledge the importance of open source hardware designs to the success of NFV implementation. This also means that the efforts we’ve made in the OPNFV community to work more closely with OCP over the past year, are paying off. In fact, OPNFV was successfully deployed on OCP hardware for the first time during our Colorado Plugfest last December held at the University of New Hampshire Interoperability Labs. To that end, OCP has a strong role in the full end-to-end open source network stack, which was discussed at the Big Communications Event in Austin as part of our OPNFV workshop earlier this year.  

The survey data also reaffirms another key aspect of OPNFV: DevOps. It’s still in the early stages, but DevOps will play a critical role in the success of NFV and this is recognized by 80% of the CSPs we surveyed, with several are either already evaluating various toolchains or working on automation and testing infrastructure. The numbers of those currently building Continuous Integration/Continuous Development (CI/CD) pipelines, or pushing out daily automated patches is lower, but recognition of the importance of DevOps and the role it will play to create a truly integrated pipeline for NFV is encouraging.

All in all, we see the steady adoption of NFV across CSP networks.  Our community is making meaningful strides as an integral player in accelerating the deployment and commercialization of NFV solutions in the market, enabled by the innovation and commitment of our growing community and our upstream integration partners. You can access the Heavy Reading survey slide presentation here, and view Roz’s session video from the Beijing OPNFV Summit here.


OPNFV Compliance Verification Program

By | Blog

By Wenjing Chu, Distinguished Engineer and Senior Director of Open Source and Standards at Huawei

In the late 80’s, I was a college student studying computer science in the northwestern suburb of Beijing, where horse driven carts could still be seen on the quiet streets and most of the city’s residents commuted via bikes and buses. The city’s transformation since then has been unprecedented and astonishing. This phenomenal development has been credited by many economists to the simple fact that ideas and products produced by one group of creative people can be used by other groups around the world with minimal friction. When a good idea becomes a common platform, it taps into an ecosystem’s power of creativity, efficiency and economies of scale. It was therefore quite fitting that the OPNFV community recently gathered in Beijing for the third-annual OPNFV Summit.

Recognizing the importance of a common platform based on open source technologies, the OPNFV Board of Directors began discussions to define a compliance testing program not long after the project’s formation. Soon after, the Dovetail project was started to implement and realize such a program by leveraging the entire OPNFV community’s efforts in many fronts, including infrastructure, testing and feature projects. The goal of OPNFV compliance testing is to reduce adoption risk of commercial OPNFV-based infrastructure by decreasing complexity and cost of testing product capabilities and interoperability. 

Formation of the CVP

In December of 2016, the Board of Directors approved the Compliance Verification Program (CVP) Guidelines as a governance document outlining strategy and the overall scope. At the Beijing Summit, the OPNFV community announced to the public   the CVP and our intent to launch the program in the second half of 2017. Through this program, products will obtain OPNFV compliance verification signifying that the product:

  • Supports key behaviors, functions, and related APIs of the OPNFV release
  • Implements certain mandatory and optional NFV functions
  • Supports interoperability between an infrastructure built on the compliant products and applications designed to run on that infrastructure
  • Is a good candidate for more extensive testing by the operator for further trials in an operator network

Verified products submitted by respective vendors are expected to differentiate themselves with extra features and capabilities, better performance, or nicer user interfaces,  but remain compliant by implementing explicitly defined interfaces, behaviors, and key features.

The First CVP Test Suite: Based on Dovetail 

As an open source community-driven project, the test scope and automation toolset will evolve over time by contributions from the active OPNFV and upstream communities, with crucial requirements and feedback from many of the world’s service providers, standardization organizations and other end users. The Dovetail project leverages the prior work from a wide range of community projects, including Openstack refstack; OPNFV Pharos; OPNFV testing projects Functest and Yardstick; OPNFV feature projects HA, IPv6, and SDNVPN; and the entire OPNFV community through their work creating and testing scenarios that produce diverse configurations. The first release of the CVP test suite is being developed based on features and interfaces found in Danube and previous releases. Currently, it covers:

  • Cloud capabilities
  • VNF lifecycle basic functions
  • Carrier network capabilities
  • Service availability

The CVP envisions that applicants can use the toolset developed in the Dovetail project to self-test or to utilize the service of third-party testing organizations that use the Dovetail toolset. These processes and guidelines are also under development by the community. 

OPNFV Delivers Common Test  Methodology

Vendors–even if their products are based on open source–all use different versions and configurations in a plethora of combinations. This diverse and complex set of differing configurations seen across the industry can be problematic and is often seen as impeding the adoption of open source based solutions in NFV. The OPNFV community is actively working with service providers to bring a common test methodology and test automation toolset, and a common reference validation set, for evaluating suitability of commercial products from service provider lab testing. This is the vision we’re working to realize through the Compliance Verification Program.

For all in the community, from vendors, to service providers, to end users, we invite you to learn more about the CVP and the Dovetail project that is driving its implementation, and invite you to contribute to this community effort. This is an excellent opportunity to help define the test cases that are most valuable to you.  A Chinese proverb says that a journey of a thousand miles starts with the first step. Today, our community is making this important first step towards our shared success in NFV.

For information on how to join this effort, please reach out to and watch the below presentation from OPNFV Summit: “Introducing the OPNFV Compliance Verification Program.”

Testing, MANO Integration, and Hardware Progress with Third OPNFV Plugfest, Hosted By Orange

By | Blog

By OPNFV Ambassador and community members, Morgan Richomme & Xiaolong Kong

The third edition of the OPNFV Plugfest, jointly organized by OPNFV and Orange, was held at the Orange Gardens, the eco-campus of Orange just outside of Paris, April 24-28.2017. The first OPNFV Plugfest to take place outside of the United States, it brought together around 100 developers and network architects from over 30 partner organizations from across Europe, Asia and North America. We were proud that Orange could host the third edition of the OPNFV plugfest end of April. As a Service provider, actively contributing to OPNFV, managing two Pharos Community Labs, it was a natural fit to host such event.

Plugfests follow the Aristotelian rules of drama: unity of action, of time and of place, and as such, are an incredible humane catalyst for the community. Being able to discuss with all the actors (installers, testing, infrastructure, features) in one place during one week prevents months of complex interoperability tests or remote discussions.

As with previous Plugfests, several activities were done in parallel: bugs were reported; solutions installed on new hardware; tests executed both on-site and on remote systems. Open Compute Project (OCP) and smart NIC activities were especially interesting, associating OPNFV with cutting-edge hardware solutions. MANO was officially introduced into OPNFV Danube (via Heat, Cloudify, Tacker, Open-O, OpenBaton). The cross-sessions with ONAP were useful to prepare OPNFV Euphrates and position OPNFV as a target infrastructure to deploy and evaluate orchestrator function, mainly through future cross CI activities.

OPNFV Plugfest is also the right place to meet people and evaluate new concepts: power consumption prototype was shared with the community, encouraging the developer to go further and leverage the Pharos Labs federation to collect more information.

Paradoxically, I think that one of the main outcomes of this Plugfest is related to the fact  that we were able to break the system. Stress tests have been introduced in OPNFV Danube and clearly identified some limitations on OPNFV, but also commercial solutions. The capability to identify such problems, to reproduce them, is crucial to evolve towards a real carrier-grade solution. In addition to destructive stress tests, the ability to set up a realistic load test campaign using a commercial virtual loader on a virtual IMS also indicated a better level of maturity of the ecosystem. 

The biggest takeaway is that Plugfest provides additional evidence that OPNFV continues to make real progress.  Despite its relatively informal atmosphere as compared to traditional and complementary interoperability events, I think the week was too short for most of the participants. We may also note Plugfests now attract a new type of actor, which includes non OPNFV members, which is fantastic.

The challenges do remain, especially related to performance, qualification and orchestration, and we certainly have our work cut out for us. But OPNFV Danube priorities clearly identified fields to be consolidated: for instance, a complete robustness/stress test strategy still needs to be built.

Learning, sharing and shaping the future: this Plugfest, like the others, was a great success!

More details on the specific outcomes of the OPNFV Danube Plugfest can be found on the official Plugfest white paper, available here.

The next OPNFV Plugfest will take place at the Intel campus near Portland, OR, December 4-8, 2017, and will focus on the upcoming OPNFV Euphrates release. Stay tuned for more details!

Many thanks to Nokia, Huawei and Advantech for having provided on-site equipment. Special thanks to Orange’s organization team, coordinated by Claire Chabrier Larroche and Xiaolong Kong, for their devoted work since the beginning of the year, without which this event could not have happened.

OPNFV Summit Keynote Preview: Q&A with HPE’s Prodip Sen

By | Blog

HPE’s Dr. Prodip Sen, CTO of NFV at Hewlett-Packard Enterprise and OPNFV Board of Directors Chair, will deliver a keynote address during the upcoming OPNFV Summit focused on “Cloudification of the Telco Network – Are we There Yet?” Read  below for a preview of what you can expect from Dr. Sen’s keynote at OPNFV Summit this year. OPNFV Summit is taking place June 12-15 in Beijing. More details, including onsite agenda and how to register, are available here:

Tell us about your involvement in OPNFV (and open source networking in general) and how it’s evolved over time. 

I am currently the Chair of the OPNFV Board of Directors, and have been the chair since the inception of OPNFV three years ago. (Time for a change I guess!) My involvement with OPNFV and open source in general started as an outgrowth of the NFV movement which some industry veterans (including myself)  helped create. We were looking for an alternative to the typical standards track to encourage faster development and deployment of the technologies required for NFV, and the telco network transformation in general. This led to my awareness of open source and the creation of the OPNFV organization. Since then, I have been engaged with activities in other open source organizations such as OpenStack and OpenDaylight.

Where do you see your role — and that of OPNFV – in terms of the broader end-to-end open networking stack?  
My interest is, and has always been, in seeing how all these open networking activities fit together and how we can make them more synergistic. I see OPNFV as a crucial organization in being the integration project that provides a space for putting the entire stack together. And eventually, showcasing use cases important to the users of this stack.

What new technology or trend in the networking space are you most curious about and why?
Not so much a trend/technology in networking per se, but I am very interested in how the industry takes the last steps of the journey to cloudify the network. We need to incorporate many of the lessons learned  from cloud-as-a-service architectures, microservices etc. and consolidate into a more constrained, quality-of-experience centered world. Will either survive the clash?

In what ways has the industry progressed towards broader adoption of NFV? What are the hurdles?
There has been a lot of progress in telco adoption over the few years since the movement was launched – from skepticism, to grudging acceptance, to “show me,” to now routine inclusion in RFIs, RFPs and trials. We all know that the telco adoption cycle is very long – so it is especially encouraging to see such a new approach gain traction in so short a time. On the other hand, the optimists among us hoped for faster adoption and true large-scale deployment by this time. The hurdles we’re facing are not new or unexpected; many in the industry hoped for (perhaps expected) a faster transformation in attitudes and organizational change, though both are still somewhat lagging.  

Can you give us a preview of what you’ll be talking about onstage at OPNFV Summit in Beijing?
I will be talking about the Cloudification and Disaggregation stages of the telco journey to the cloud, and what we in the industry, and open source and OPNFV in particular, need to pay attention to.

OPNFV Summit Keynote Preview: Q&A with ZTE

By | Blog

ZTE’s Zhang Fan, Chief Architect of Packet Core, will deliver a keynote address at the upcoming OPNFV Summit focused on “NFV Practice for vEPC Commercial Network.” Read  below for a preview of what you can expect from ZTE onsite at OPNFV Summit this year. OPNFV Summit is taking place June 12-15 in Beijing. More details, including onsite agenda and how to register, are available here: 

Tell us about your involvement in OPNFV — and open source networking in general. How has this evolved over time?
ZTE’s open source road began in 2014. Now, more than 200 professional developers from ZTE are working with partners in 13 open source communities. Recent data shows ZTE is among the top contributors within the OPNFV, OpenStack, and kubernetes communities. ZTE looks forward to continued work with SDN/NFV ecosystem partners to promote maturity of open source and industrial software development.

More specifically, ZTE has released enhanced network solutions, including an SDN controller, SDN switches, intelligent NICs and data plane acceleration. ZTE ZENIC SDN Controller, ZTE DVS (OVS based) and mobile core network are using these technologies via open source.

ZTE’s involvement in open source is currently employed within several commercial products and projects like ZX-TECS based on Openstack and ZX-vManager based on OPEN-O.

Where do you see your role — and that of OPNFV — in terms of the broader end-to-end open networking stack?  
OPNFV plays a key role in the integration of IT technologies and standards organizations by centralizing the ecosystem via a dedicated reference platform, which speeds up NFV development and maturity. Since OPNFV representatives are comprised of CT vendors, operators, and IT vendors, among others; OPNFV reflects common requirements needed across the NFV ecosystem. For example, OPNFV’s strong collaboration with other upstream communities including OpenStack, OpenDaylight, DPDK,, etc. illustrate the project’s ability to serve as a connection across the end-to-end open networking stack.  

OPNFV is the only open source community targeting  NFV solutions, including infrastructure, VIM and MANO. ZTE is pleased to be joining the OPNFV community along the journey to accelerate open source NFV.

What new technology or trend in the networking space are you most curious about and why?
As NFV/SDN technologies have been deployed by operators and will continue to enable 5G and IoT applications, network infrastructure must be optimized to adjust to a centralized management for controlling the plane, high throughput traffic and ultra-low latency for the data plane. Cloud deployment needs to be extended from the core to the edge. The focus of networking is shifting from functions to efficiency and performance.

On the one hand, SDN is playing a critical role in coordinating all open source components to optimize the network architecture and improve network efficiency. SDN makes it possible to manage the entire network through intelligent orchestration and management systems in a 5G network. So it is quite interesting to investigate how best to provide continuous delivery (CD) of services based on dynamic network functions and automated deployment mechanisms in future networks.

On the other hand, the performance must be improved to meet the requirements of 5G and AR/VR, etc. Currently, we already focus on the software level to identify mechanisms to enhance the processing as fast as possible. By introducing upstream projects like, we can achieve more data plane acceleration. On top of that, hardware acceleration is also an up-and-coming technical trend. We need to take into consideration that some complex computing or algorithm processing can be offloaded to intelligent cards, which will release a lot of hardware core resources of the data plane being used.

In what ways has the industry progressed towards broader adoption of NFV? What are the hurdles?
For starters, the proposed NFV architecture or solution must provide a cost-effective way to help operators build their networks. There are many open source communities; we need to integrate them together to make networking simpler and more agile. Currently, a lot of operators have started transforming their network, so it is the right time to introduce NFV to replace traditional network components.

Secondly, the industry should adopt more and more vendors and apps to build an open ecosystem instead of the traditional “chimney style.” We should consider how to provide a common platform to make all components integrate more closely. For example, ZTE ElasticNet integrates SDN and NFV technology in order to break the isolation between components. This gives customers a better experience in terms of management, service and new business models.

Additionally, the industry needs to target upcoming new networks and new services. The architecture must be adapted to future (5~10 years’) requirements. With 5G and IoT booming, cloud native- based technology is a good way to broaden NFV.

One of the major hurdles towards broader adoption of NFV is closed architecture and deployment methods. Some industry projects are not authentic NFV networks, but just simple virtualization within one vendor. We need to break vendor lock by leveraging open cloud networks, where customers can select  products based on different layers and needs;  otherwise, big barriers continue to block large-scale NFV application growth.

Can you give us a preview of what you’ll be talking about onstage at OPNFV Summit in Beijing?
We’re looking forward to participating in this year’s OPNFV Summit! In addition to introducing commercial NFV mobile network practices beneficial to open source community development during the main keynotes, ZTE will be onsite for the Design Summit beginning Monday, June 12. We are excited to collaborate with the community, particularly regarding plans for the next OPNFV release, Euphrates.