January 2017

Luis Jorge Romero, ETSI Director General

When ETSI created its first Open Source group last year, many in the industry were surprised. Open Source MANO was ETSI’s first real step into Open Source, although the subject has long been discussed. I would say it was a necessary step for ETSI.

In the past, Open Source was somehow seen as a rival to standards. It wasn’t just that the economic and licensing models were different, the working methods and makeup of the communities were also quite different. Today, however, Open Source is broadly accepted as having its place in almost all areas of software development. Our standards are increasingly implemented in software. Our industry members have embraced Open Source. ETSI needs to evolve in pace with its members.

ETSI’s future lies in our ability to attract new communities and evolve to meet the needs of new members. This includes learning from the experience of others and adapting our ways of working. The world of Open Source has developed tools and processes to aid collaborative development of software by remote developers. Standards development often faces similar problems: building and attracting communities, ensuring quality of work, completeness and meeting time to market demands. We must be prepared to redesign our ways of working to take account of best practice, wherever it may lie.

In the case of Open Source MANO, we needed quickly to learn new ways. We have had to be flexible and adapt our procedures, our support levels and our IT infrastructure to meet the expectations of this new community. We have adopted all the Open Source tools with which the community is used to working. We’ve been willing to push for the success of this group. From the feedback they give me, I think we have succeeded.

ETSI is serious in this new approach. We anticipate that industry may want to follow with other Open Source initiatives in the wake of Open Source MANO. If this is the case, then we’re ready!

Speed matters. In software development, no matter whether it is proprietary or open source, early success is often defined in the same way as an Olympic 100-meter dash. The first few steps out of the starting block are critical in quickly getting to the 100m point. The race then changes once this has been accomplished – the straight line race track starts to curve and a group of runners settles in to run the distance and find success at the finish line. 

To me, the first year of OSM bears a number of similarities to a 100-meter race, and as the year 2017 has just started, I am taking some time to reflect on what we have jointly accomplished since the inception of the OSM community and outline a number of personal thoughts on key themes for this new year.

Not to wax nostalgic, but looking back at 2016 does showcase a number of accomplishments that would be the envy of the majority of open source communities. From the first public OSM demonstration at the end of February at MWC 2016, to creating two OSM releases that have been downloaded more than 2000 times from 50+ countries, to capturing the interest of 55 organizations that have joined OSM, we have much to be proud of. On the technical side, the creation of a well-functioning, one-step installable, multi-VIM, multi-SDN controller OSM Release ONE that incorporates information models commonly agreed upon by a group of operators and solution vendors stands out to me. On the community engagement side, the OSM workshop at SDN World Congress in The Hague, the launch of the new website, quarterly newsletters, the OSM Twitter channel and, personally most important, the contributions of an awesome group of marketing professionals to create awareness and excitement for OSM top my list.

Year 2 will be different. Speed still matters, but it will be accompanied by well-measured collaboration activities to take the OSM software stack through the standard lab/field/production deployment cycle with service providers engaging in MANO. Interoperability, stability, security and the size and maturity of the commercial support ecosystem for OSM will become key themes on the road to production deployments. New use cases and network service scenarios from the broader community will tie the code base to its economic value. I expect that OSM’s guiding principle of modularity for any software component will provide additional choice and capabilities. 

In addition, as a result of the success of broadly observed initiatives such as Telefonica’s UNICA program, a set of existing and new community members will choose to actively invest time and contribute engineering expertise and code as purchase points for OSM proliferate across the globe. ETSI’s groups for NFV and MANO will receive an increasing amount of input based on the pragmatic results of the work within OSM.

So speed and timing will still matter as OSM turns its sights from a 100-meter dash to a 10k race in 2017. I believe that as with other successful software development initiatives, collaborative and reliable execution and a singular focus on unlocking OSM’s economic value for the entire community over the next 5-10 years will set us apart. OSM’s accomplishment in getting out of the starting block with speed and clear direction is the best indicator for continued success in 2017 and beyond.

At last year’s SDN World Congress in Dusseldorf the concept of an open source NFV MANO platform was just an idea. 12 months later it’s trial-ready code. The ETSI Open Source MANO (OSM) community kicked off SDN World Congress with a technically deep and, at times, entertaining OSM workshop. The workshop featured learning sessions from each of the seed code contributors, RIFT.io, Telefonica, and Canonical as well as terrific insights from the OSM Technical Steering committee and End User Advisory Group. The buzz surrounding OSM was evidenced by the nearly 100 attendees that came to learn more about the recent OSM Release ONE and OSM’s development roadmap. 

OSM’s Chairman, Francisco-Javier Ramón, got things started with an introduction to the OSM community and OSM Release ONE. Francisco Javier covered the key features of Release ONE and Release ONE’s goal of making it easy to test OSM and build trials. He also introduced OSM’s network of remote labs that enables continuous and automated testing with different types of infrastructures (VIMs, SDN Controllers and NFVIs). OSM has made some bold plans for the next six months, including the support of new types of infrastructures, integrated approaches to Security and Service Assurance, the participation in the upcoming ETSI NFV Plugtest™, and new POCs and demos based on Release ONE. The OSM community is making a bold statement that Release TWO will be deployment ready. 

Read more ...