Doctor of Philosophy (PhD)


Computer Science and Engineering

Document Type



Cloud storage has gained increasing popularity in the past few years. In cloud storage, data are stored in the service provider’s data centers; users access data via the network and pay the fees based on the service usage. For such a new storage model, our prior wisdom and optimization schemes on conventional storage may not remain valid nor applicable to the emerging cloud storage.

In this dissertation, we focus on understanding and optimizing the I/O performance and cost efficiency of cloud storage from a client’s perspective. We first conduct a comprehensive study to gain insight into the I/O performance behaviors of cloud storage from the client side. Through extensive experiments, we have obtained several critical findings and useful implications for system optimization. We then design a client cache framework, called Pacaca, to further improve end-to-end performance of cloud storage. Pacaca seamlessly integrates parallelized prefetching and cost-aware caching by utilizing the parallelism potential and object correlations of cloud storage. In addition to improving system performance, we have also made efforts to reduce the monetary cost of using cloud storage services by proposing a latency- and cost-aware client caching scheme, called GDS-LC, which can achieve two optimization goals for using cloud storage services: low access latency and low monetary cost. Our experimental results show that our proposed client-side solutions significantly outperform traditional methods. Our study contributes to inspiring the community to reconsider system optimization methods in the cloud environment, especially for the purpose of integrating cloud storage into the current storage stack as a primary storage layer.



Committee Chair

Chen, Feng