论文部分内容阅读
由于应用需求的快速发展以及网络存储系统的出现,因此异构磁盘阵列的变得越来越常见。RAID5由于较高的性能和可靠性以及较低的代价,是应用最为广泛的RAID结构。目前对异构磁盘阵列RAID5结构的研究,重点主要放在充分利磁盘存储空间以及对性能的定性研究。论文提出了一种异构磁盘阵列RAID5结构数据布局优化方法,该方法充分考虑异构磁盘的相对容量和性能,以及校验单元的散布对RAID5小数据写性能的影响,可以生成负载均匀分布或接近均匀分布的布局。仿真实验结果表明,对于多用户小数据访问模式,优化布局的性能明显优于简单RAID5布局,且具有更高的伸缩性。
Heterogeneous disk arrays are becoming more common due to the rapid growth of application requirements and the advent of networked storage systems. RAID5 is the most widely used RAID fabric due to its high performance, reliability and low cost. At present, the research on RAID5 structure of heterogeneous disk array mainly focuses on the sufficient disk storage space and the qualitative research of performance. In this paper, we propose a data layout optimization method for heterogeneous disk array RAID5 architecture. The method fully considers the relative capacity and performance of heterogeneous disks and the impact of the spreading of parity units on the performance of RAID5 small data writes. It can generate uniform load distribution or Close to evenly distributed layout. Simulation results show that for multi-user small data access mode, the performance of the optimized layout is obviously better than the simple RAID5 layout, and has higher scalability.