Enhancing the Performance of an Autonomous Distributed Storage System in a Large-Scale Network
Atsushi Nunome, Hiroaki Hirata · 2022
We are developing an autonomous distributed storage system that autonomously relocates data blocks according to their utilization and the effective performance of storage. This paper proposes an enhancement to the distributed storage system to address large network environments where many network nodes are connected to multiple network switches. In the proposed method, the latency time when a network frame travels on the network is evaluated together with the effective performance of the storage, and this is used to determine the destination of the data block. Simulation evaluation results show that in the networks with cascading multiple network switches and high I/O load on storage, the performance improvement is significantly greater than that of the previous method. This showed that our storage system can provide sufficient performance improvement even in large network environments.