Cloud providers now often provide a new type of server abstraction called transient servers. Transient servers offer significantly lower costs than their on-demand equivalents with the added complication that the cloud provider may revoke them at anytime. Transient servers’ low price makes them particularly attractive for executing application workloads that are resource-intensive. 

In this talk, we illustrate the challenges and opportunities of executing various cloud workloads on transient servers, where revocations (akin to failures) are common cases.

Tian Guo is an Assistant Professor in the Computer Science Department and Data Science Program at Worcester Polytechnic Institute. Her research interests include distributed systems, cloud computing, mobile computing and cloud-enabled IoTs. Her current focus is on handling dynamics introduced by new cloud workloads, e.g., training and hosting deep learning models, and emerging cloud platforms. She received her Ph.D. and M.S. in Computer Science from the University of Massachusetts Amherst in 2013 and 2016, respectively, and her B.E. in Software Engineering from Nanjing University in 2010. Her work is supported in part by the NSF CRII award and Google Cloud Research credits.