Optimizing performance and reliability in distributed computing systems through wide spectrum storage

In this paper, we provide an overview of logistical runtime system (LoRS). LoRS is an integrated ensemble of tools and services that aggregate primitive (best effort, faulty) storage allocations to obtain strong properties such as capacity, performance, reliability, that Grid applications desire. The paper focuses on the design and implementation of LoRS, as well as the storage scheduling decisions that LoRS must make.