|Large corporations, such as Google and Facebook, rely on erasure coding to reduce the storage cost for a given reliability. In modern storage systems, the rapid growth of data traffic, such as streaming, e-commerce, etc., has stressed underlying data storage systems. Historically, a key solution to relieve this traffic burden is caching. Chunks of popular data can be stored separately closer to end-users, reducing congestion in the network and improving delay times. Among stored data, 20 percent of the content may be accessed 80 percent of the time, so establishing a priority on this data improves functionality significantly.
Researchers at Purdue University, George Washington University, and AT&T Research have developed a novel caching framework with erasure code called functional caching, which uses erasure coded chunks in the cache. An algorithm optimizes caching and quantifies an upper bound on the mean service latency in closed-form for storage systems. Arbitrary chunk placement and service time distributions are formulated via the efficient heuristic algorithm. A simulation prototype, operated in an open-source cloud storage system, validates significant service latency reduction utilizing this caching method.
-Monitors placement and distribution
-Data storage system
-Erasure coded caching
Jun 30, 2016
Purdue Office of Technology Commercialization
1801 Newman Road
West Lafayette, IN 47906
Phone: (765) 588-3475
Fax: (765) 463-3486