Predictable Data Centers


July 23, 2015


Performance predictability is a key requirement for high-performant applications in today’s multi-tenant massive-scale data centers hosting hundreds of thousands of servers. Online services like search or social networks running in infrastructure data centers need such predictability to satisfy application Service Level Agreements (SLAs); for example, online search queries need to be responded within a few hundred milliseconds. Cloud data centers require guaranteed performance to bound customer costs and spur adoption. However, several components of today’s datacenters are at odds with such high-level application SLAs.

The Predictable Data Centers (PDC) project tackles the issue of unpredictable application performance in data centers. A key contributor to such unpredictability is shared resources like network and storage. The bandwidth across the cloud network and to the cloud storage service can vary significantly. In this talk, I will discuss how we can design a predictable data center architecture that offers performance SLAs across shared resources


Thomas Karagiannis

Thomas Karagiannis is a researcher with the Systems and Networking group of Microsoft Research Cambridge, UK since 2006. He received his Ph.D. in Computer Science from the University of California, Riverside and his B.S at the Applied Informatics department of the University of Macedonia, in Thessaloniki, Greece. Before joining Microsoft, he has also been with Intel Research and the Cooperative Association for Internet Data Analysis (CAIDA). Thomas received the 2011 ACM SIGCOMM Honorable Mention award, and has several publications in the top conferences for computer networks. His research interests include data center networks, Internet measurements and modelling, and social networks.