Distributed storage can and should be elastic, just like other aspects of cloud computing. When storage is provided via single-purpose storage devices or servers, separated from compute activities, elasticity is useful for reducing energy usage, allowing temporarily unneeded storage components to be powered down. However, for storage provided via multi-purpose servers (e.g. when a server operates as both a storage node in a distributed filesystem and a compute node), such elasticity is even more valuable— providing cloud infrastructures with the freedom to use such servers for other purposes, as tenant demands and priorities dictate. This freedom may be particularly important for increasingly prevalent dataintensive computing activities (e.g., data analytics).
This project develops novel designs for elastic storage capable of adapting rapidly to I/O workload intensity, including our groundbreaking SpringFS and Rabbit systems.
FACULTY
GRAD STUDENTS
James Cipar
Nitin Gupta
Elie Krevat
Alexey Tumanov
Lianghong Xu
INDUSTRY COLLABORATORS
Michael A. Kozuch, Intel
This research was funded (in part) by the Intel Science and Technology Center for Cloud Computing.
We thank the members and companies of the PDL Consortium: Amazon, Datadog, Google, Honda, Intel Corporation, IBM, Jane Street, Meta, Microsoft Research, Oracle Corporation, Pure Storage, Salesforce, Samsung Semiconductor Inc., Two Sigma, and Western Digital for their interest, insights, feedback, and support.