Jennifer Rexford is not the type of thought-leader who offers grand statements about her ideas for remaking the status quo. Her answers to questions are succinct, knowledgeable and have a touch of humor. Her visions are spun around code and through her work with students. She puts her head down and builds the tools she hopes will change the networking world, and open it up to innovation for the many as opposed to the few.
Rexford, who is the Gordon Y. S. Wu Professor in Engineering at Princeton, has gained a following in the networking world for her work building Frenetic, a programming language for software-defined networks. Her interest in networks stems from her love of math and a thin streak of rebellion. It’s clear that she’s not interested in authority for the sake of authority, whether that’s Cisco or Juniper dominating how people program networks or the “thugs in airport security,” that she mentions in a joke during our interview.
Cisco, let my networks go
You can also hear that streak of rebellion in her story about how she started down the path to computer science and networking back when she was a high school sophomore in the mid 1980s and got her first computer.
“This somewhat relates a to my interest in SDN, but the thing I found most interesting was that someone got to decide what instructions the computer gets to run,” Rexford said. “And I thought that was super nifty and very empowering, so this idea of lowering the barrier, of deciding who gets to innovate and who gets to design the instruction set for the computer, which software it runs … it’s how you design things so you decide who gets to innovate. And with SDN you can allow a much wider range of people to innovate on top of the network.”
Her work centers around creating a modular programming language around networking that allows more people to build networking applications to fit their needs. The Frenetic programming language breaks up the monolithic network services, from routing and monitoring to security and load balancing, into units that programmers can control and play with in building other apps. So if you want to tie load balancing into some aspect of your applications, you don’t have to drag these other elements along, and can pick and choose what you use.
Like Martin Casado, the co-founder of Nicira and another SDN expert, Rexford didn’t start out thinking about software-defined networks. She instead had majored in electrical engineering and spent her graduate and doctoral school time studying massively parallel and distributed computing. Unfortunately, once she graduated with her Ph.D., all of the parallel computing startups had gone bankrupt, which left her at somewhat of a loss.
Everything is distributed, so now what?
But the growth of the internet and her interest in distributed systems meant networking was a good alternative, so Rexford started focusing on that work. And despite almost nine years working at AT&T, Rexford has spent her time since in academia, teaching and completing research related to networking. She contributes to the Open Flow community and participates in the Open Networking Summits. That outreach and commitment to opening up the networking world is why we’ve named her a Cloud Trailblazer this year.
Guru Parulkar, the chair of the Open Networking Summit notes that not only is Rexford seemingly always online with invaluable information to share, she’s an incredibly positive person. “I haven’t seen her to be critical of any piece of work or person in a negative way,” he said.
That positivity might explain one of Rexford’s few activities outside of her research. She offers IT help to non-profit organizations, especially those in the LBGT community, where she and her partner are both active. She also offers that she’s into food and wine: “Gluttony is my sin of choice.” But Rexord, who took less than a minute to share her life story and was eager to get back to her meetings with students and her work, appears to be hesitant to discuss introspection or share idle chatter.
However, if you get her talking about inter-domain routing, economic incentives behind the internet or the future of software-defined networks on security, she visibly sparks. She then switches into coherent and paragraph-length explainers on how the internet works and where the weaknesses lie. Much of it parallels her themes around opening things up to innovation by the many and emphasizes her collaborative style of working with other people, government agencies and possibly anyone who wants to participate. Like her work, Rexford is open to innovation and ideas from anyone with a good one.