Adaptive Parameter Tuning for Constructing Storage Tiers in an Autonomous Distributed Storage System

Abstract We are developing an autonomous distributed storage system to make effective use of networked heterogeneous devices. It dynamically configures the tiers using performance values measured during operation, rather than static indicators device. Data blocks transferred a more appropriate tier according their read/write tendencies. In previous paper, were set statically defined thresholds, but in this we propose method thresholds adaptively. This eliminates need adjust runtime parameters for threshold setting environment. has also been improved so that parameter can be individual storage, applying single value entire system. system, administrator had readjust periodically match usage status device and behavior process. However, proposed method, fine-tuned operation reduce workload administrator. The simulation results show execution time by up about 30% compared with optimum configured statically.