Optimizing Probe-Based Storage

Probe-based storage, also known as micro-electric mechanical systems (MEMS) storage, is a new technology that is emerging to bypass the fundamental limitations of disk drives. The design space of such devices is particularly interesting because we can architect these devices to different design points, each with different performance characteristics. This makes it more difficult to understand how to use probe-based storage in a system. Although researchers have modeled access times and simulated performance of workloads, such simulations are time-intensive and make it difficult to exhaustively search the parameter space for optimal configurations. To address this problem, we have created a parameterized analytical model that computes the average request latency of a probe-based storage device. Our error compared to a simulated device using real-world traces is small (less than 15% for service time). With this model we can identify configurations that will satisfy specific performance objectives, greatly narrowing the search space of configurations one must simulate.