Fb Fellow Highlight: Pioneering self-monitoring networks at scale
Every year, PhD students from around the world apply for the Facebook Fellowship, a program designed to encourage and support promising PhD students doing innovative and relevant research in computer science and engineering at an accredited university.
As a continuation of our Fellowship Spotlight series, we’re highlighting Facebook Fellow in 2020 Networking and connectivity, Yaseen’s novel.
Nofel is a fourth year PhD student at the University of Pennsylvania, advised by Vincent Liu. Nofel’s research interests lie in a number of related topics in the broad areas of distributed systems and networking. His most recent work aims to develop a fine-grained network measurement tool and see how we can use these measurements to understand network traffic patterns.
Nofel is fascinated by the Internet and how it enables us to communicate worldwide, from anywhere and at any time. “The sheer complexity of the network system used to keep the world online is what fascinates me to understand better,” he says.
This led Nofel to pursue a PhD in Network Systems, where he was able to study the many types of devices used in different systems – servers, switches, and network functions. His research posed a question: how do we currently monitor networks and is there any way we can monitor them more effectively?
“Networks keep getting bigger, and as they expand we need more innovation to handle the size and growth of the network,” explains Nofel. As networks grow, how to monitor network status and performance becomes more complex. Humans can only monitor a limited amount, and the reality of expanding networks requires a tool that can automate network monitoring. These questions led him and his staff to develop a tool called Speedlight, a fine-grained network measurement tool that can help us better understand network traffic patterns.
Speedlight has been deployed and tested in small topologies, but Nofel hopes to scale his work on Speedlight to meet the needs of larger networks. “Networks are constantly evolving and require more research and innovation to come up with new solutions that can handle huge networks,” he says, and network monitoring scaling tools must address provisioning and traffic challenges. Through the Facebook Research Fellowship, Nofel has reached out to Facebook engineers to understand the industry problems larger data centers are facing in terms of surveillance so that he can focus his work more intensely.
“I asked myself, ‘How can networks be self-driving?’” He says. “How can you monitor and debug yourself? How do operators choose what they want to provide in their own data centers? How do they decide what to run and what not? ”Nofel looks forward to further research and exploration of how networks can incorporate machine learning and deep learning into further developments in monitoring, as they are a natural extension of the development of self-monitoring ones Networks are.
To learn more about Nofel Yaseen and his research, visit his website.