Application developers using a cloud storage system often make decisions that trade off consistency for performance and may be locked into a choice that is not ideal for all clients. Pileus is a replicated key-value store that allows applications to declare their consistency and latency priorities via consistency-based service level agreements (SLAs). It dynamically selects which servers to access in order to deliver the best service given the current configuration and system conditions. In application-specific SLAs, developers can request both strong and eventual consistency as well as intermediate guarantees such as read-my-writes, monotonic reads, bounded staleness, and causal consistency. Evaluations running on a worldwide test bed with geo-replicated data show that Pileus adapts to varying client-server latencies to best meet a given consistency-based SLA.
There is no downloadable version of the slides for this talk available at this time.
About the speaker:
Doug Terry is a Principal Researcher in the Microsoft Research Silicon Valley Lab. His research focuses on novel distributed systems, including cloud data services and mobile computing. Prior to joining Microsoft, Doug was the co-founder and CTO of a start-up company named Cogenia, Chief Scientist of the Computer Science Laboratory at Xerox PARC, and an Adjunct Professor in the Computer Science Division at U. C. Berkeley, where he regularly taught a graduate course on distributed systems. He is Chair of the ACM Tech Pack on Cloud Computing, the past Chair of ACM’s Special Interest Group on Operating Systems (SIGOPS), and an ACM Fellow. Doug received a Ph.D. in Computer Science from U. C. Berkeley where he helped develop Berkeley UNIX.
1065 La Avenida
Mountain View, CA 94043