Beyond Binary Failures: A New Model for Networks

Monday, February 19, 2018 -
4:00pm to 5:00pm
CS 1240

Speaker Name: 

Monia Ghobadi

Speaker Institution: 

Microsoft Research Mobility and Networking

Cookies: 

Yes

Cookies Location: 

CS 1240

Description: 

Abstract:
Fiber optic cables are the workhorses of today's Internet services. They are an expensive resource acquired by organizations with significant monetary investment. Their importance has driven a conservative deployment approach with redundancy baked into multiple layers of the network with the assumption that links have a constant reliability status and operate at a fixed capacity. In this talk, I will take an unconventional approach and argue that link failures should not be always considered as binary events enabling the foundation of a model of networks with links that have a dynamic capacity and reliability levels. I will investigate this idea by conducting the first large-scale study of operational optical signals analyzing over 2,000 channels in a wide-area network for a period of 3 years as well as 350,000 links in 20 data center networks worldwide. My analysis uncovers several findings that enable cross-layer optimizations and smart algorithms to improve traffic engineering and network planning, increase capacity, and reduce cost. For instance, I will show that the capacity of 99% of wide-area links can be augmented by at least 50 Gbps, leading to an overall capacity gain of more than 100 Tbps. This means using the same links, we get higher capacity and better availability. In data center networks, I will show that 99.99% of the links have an incoming optical power level that is higher than the design threshold and by allowing links to have multiple reliability levels, we can cut the cost of data center networks by nearly half. Moreover, this model opens the door to revisit several classical networking problems, such as the maximum-flow problem and graph abstractions, in light of dynamic capacity and reliability links and I will provide the foundation to solve these problems. In addition to new avenues of research, Microsoft has invested in this new model and is rolling out the necessary infrastructure for deployment.

Bio:
Monia Ghobadi is a researcher at Microsoft Research Mobility and Networking research group. Prior to MSR, she was a software engineer at Google. She received her Ph.D. in Computer Science at the University of Toronto and B.Eng. in Computer Engineering at the Sharif University of Technology. Monia is a computer systems researcher with a networking focus and has worked on a broad set of topics, including data center networking, optical networks, transport protocols, network measurement, and hardware-software co-design. Many of the technologies that she has helped develop are part of real-world systems at Microsoft and Google. Monia has been recognized as the N2women rising stars in networking and communications in 2017. Her papers have won the best dataset award, Google research excellent paper award (twice), and the ACM IMC best paper award.