Benjamin Cassell

Learn More
In this paper, we present Nessie, a low-latency cuckoo hash table design that uses only one-sided RDMA operations to perform read and write requests. Nessie makes use of a self-verifying data-structure to handle reads that occur in parallel to writes, and atomic RDMA compare-and-swap operations to apply multiple operations with at most one data modification(More)
The increasing use of key-value storage systems in performance-critical data centre applications has motivated new storage system designs that use Remote Direct Memory Access (RDMA) to reduce communication overhead. However, existing approaches that achieve low latency and high throughput do so by dedicating entire cores to RDMA message handling, including(More)
To my wife Shetia and my son Kaden ii ACKNOWLEDGEMENTS First off, I would like to thank God for putting me in the position that I am in today. As any Ph.D student knows, there are many times where the academic grind and struggle of completing a doctoral degree seems overwhelming. Thus, I have to thank God for putting the right people in my life to keep me(More)
—Most cloud providers improve resource utilization by having multiple tenants share the same resources. However, this comes at the cost of reduced isolation between tenants, which can lead to inconsistent and unpredictable performance. This performance variability is a significant impediment for tenants running services with strict latency deadlines.(More)
—In this paper, we introduce Mayflower, a new distributed filesystem that is co-designed from the ground up to work together with a network control plane. In addition to the standard distributed filesystem components, Mayflower includes a flow monitor and manager running alongside a software-defined networking controller. This tight coupling with the(More)
Past work has shown that disk prefetching can be an effective technique for improving the performance of disk bound workloads. However, the performance gains are highly dependent on selecting a prefetch size that is appropriate for a specific system and workload. Using a prefetch size that is too small can lead to poor overall disk throughput, whereas(More)
  • 1