Answers are grounded in this paper and cite their sources.
Amazon.com has introduced the Simple Storage Service (S3), a commodity-priced storage utility. S3 aims to provide storage as a low-cost, highly available service, with a simple 'pay-as-you-go' charging model. This article makes three contributions. First, we evaluate S3's ability to provide storage support to large-scale science projects from a cost, availability, and performance perspective. Second, we identify a set of additional functionalities that storage services targeting data-intensive science applications should support. Third, we propose unbundling the success metrics for storage utility performance as a solution, to reduce storage costs.
No takes yet. Share an insight, caveat, or question.
Palankar et al. (2008) studied this question.
Synapse has enriched one closely related paper. Consider it for comparative context: