Benchmarking Large-scale Object Storage Servers

The power required to maintain a massive storage infrastructure is impractically high. Although research on the development of energy efficient storage systems has been conducted, it is still insufficient for easily comparing their performances. In this paper, we present a benchmark framework for large-scale object storage servers that is intended at facilitating performance comparisons of storage servers in terms of latency and throughput. A key design goal is extensibility so that the framework can accommodate commercial workload. For this purpose, we extend the Yahoo! Cloud Serving Benchmark (YCSB) framework such that it can interact with a Ceph object cluster and runs YCSB standard workloads on a Ceph infrastructure test bed. We report the results of insert, read, update, and delete operations in the Ceph infrastructure test bed.This research was supported by the MSIP (Ministry of Science, ICT and Future Planning), Korea and NAVER Corp., under ICT/SW Creative research program supervised by the IITP (Institute for Information & communication Technology Promotion) (IITP-2015-R2212-15-0025), and in part by Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIP) (No. B0101-15-0557, Resilient Cyber-Physical Systems Research). The corresponding author is K. Kang