& cplSiteName &

Tracking Telco Progress in Open Networking

Roz Roseboro

I recently attended the Open Infrastructure Summit (formerly the OpenStack Summit -- more on this later). While the conference has gotten smaller, I didn't hear any complaints. Instead, people attending recognized that the developers who showed up came to work on real problems, rather than to cheerlead for the latest technology. Despite some of the dire headlines of the past year, I did not sense a crisis or panic.

The Kubernetes vs. OpenStack issue hardly was mentioned, in public or private conversations (short version: they will both co-exist with each doing what it does best. End of discussion). Rather, to a person, the focus is on making OpenStack even better (more Day 2 functionality please!) and bringing more people under the open infrastructure tent. Note that I used the phrase "open infrastructure" rather than "OpenStack" in the second part of the sentence. Indeed, as the conference title makes clear, OpenStack does not exist in a vacuum, and the community, from leadership on down, is making collaboration with other projects a high priority.

Instead of re-hashing what you can read on OpenStack's site, I'd like to share my key takeaways from conversations with people knee deep into OpenStack. On the vendor side, I spoke with Sandro Mazziotta, director of NFV product management at Red Hat, and Christopher Price, president of Ericsson Software Technology. On the provider side, I spoke with Mohammed Naser, CEO of Vexxhost (a public cloud provider), Beth Cohen, Distinguished Member Technical Staff, NFV/SDN Network Product Strategy, at Verizon and Amy Wheelus, vice president, network cloud, at AT&T. To round things out, I also met with Jonathan Bryce, executive director of the OpenStack Foundation (OSF), and Mark Collier, chief operating officer at OSF. I asked each of them about what's working, what could work better, and what they'd like to see happen in the future.

Widespread benefits today…
To loosen up my interviewees, I lobbed what I thought was a softball question: what is your proudest moment or accomplishment from the past year? I expected to get answers around OpenStack development or community building. Instead, from both the vendors and service providers, I heard about commercial successes. This alone is notable since it demonstrates OpenStack's tangible and quantifiable value. For Verizon and AT&T, it was about how quickly they were able to launch new services. For Vexxhost, it was about moving up the ranks of public cloud providers in the OpenStack space. Ericsson and Red Hat both cited customer wins. The OpenStack Foundation's response, not surprisingly, was around something internal, but no less significant: formalizing and finalizing a new project structure.

All three service providers articulated how working with open source saves them time in developing and launching new services and allows them to benefit from the innovation of a large ecosystem. What I found interesting is how forceful they were about getting more people engaged with open source. I had expected that from the OpenStack Foundation folks, but everyone else was just as passionate about growing the ecosystem. The time for questioning if open source will work for service providers and network operators is over, in their view, and all those people who are consuming open source in the shadows should start contributing so the ecosystem can innovate even more rapidly. To me, the most intriguing insight came from Jonathan Bryce. He said that companies who commit to adopting OpenStack/open source will find that doing so changes their culture. It forces them to be more flexible and take a more external view that will help combat insularity and react better in a dynamic environment.

…but still more to be done
This is not to suggest that all is rosy in OpenStack and in open source more broadly. No one denies that OpenStack can still be too big or too unwieldly for some existing use cases -- let alone for the distributed architecture scenarios that are quickly becoming a reality. Networking, especially with multicloud, remains a particular concern. Another is how quickly network function owners are re-architecting them to be more cloud-native. This brings up a whole different set of issues, including the ROI for the function owners, given that re-architecting represents an enormous cost for them, but one that is part of the discussion since the availability of cloud-native network functions will drive demand for cloud infrastructure, including OpenStack.

Better together
The single most common complaint -- or opportunity -- is for the many (many!) open source projects addressing similar areas, like "the edge" (and let's not go into the many definitions of what "the edge" is), to work more collaboratively with each other. For what it's worth, I heard this frequently at the Open Networking Summit in April as well. My research also suggests that network operators are overwhelmed by choice -- to the point that some are paralyzed, afraid of betting on the wrong horse. To their credit, the OpenStack Foundation brought numerous representatives from other projects to the conference to facilitate cross-project dialogue, but there is a clear sense that much more needs to be done to minimize competition and duplicate work. There are only so many projects a given community can support, and the sooner clarity and agreement can be reached on who should do what, the sooner the community will benefit. Along with each person I met with, my one wish for the open source community is to please try to get along. That said, there are so many smart, dedicated people trying to make open source work for the telecommunications industry that I'm optimistic that instead of celebrating amazing technologies, we'll soon be celebrating amazing outcomes.

— Roz Roseboro, Principal Analyst, Cloud Infrastructure & Management, Heavy Reading

(0)  | 
Comment  | 
Print  | 
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
More Blogs from Heavy Lifting Analyst Notes
The Super C-band development from Huawei adds an additional 50% to a standard C-band implementation.
AI and machine learning are already key components in the closed-loop control of BT's programmable network.
Automation offers the best hope for managing today's vastly complex networks – so what's the hold-up?
Featured Video
Flash Poll
Upcoming Live Events
September 17-19, 2019, Dallas, Texas
October 1, 2019, New Orleans, Louisiana
October 2-22, 2019, Los Angeles, CA
October 10, 2019, New York, New York
November 5, 2019, London, England
November 7, 2019, London, UK
December 3-5, 2019, Vienna, Austria
December 3, 2019, New York, New York
All Upcoming Live Events