摘要：

对于同类型的I/O请求，基于闪存固态盘的请求响应时间与请求大小基本呈线性比例关系，并且固态盘的读写性能具有非对称性。针对该特性，提出一种基于请求大小的固态盘I/O调度(SIOS)算法，从I/O请求平均响应时间的角度提高固态盘设备的I/O性能。根据读写性能的非对称性，对读写请求进行分组并且优先处理读请求。在此基础上首先处理等待队列中的小请求，从而减少队列中请求的平均等待时间。采用SLC和MLC 2种类型的固态盘进行实验，在5种测试负载的驱动下与Linux系统中的3种调度算法进行比较，对于SLC固态盘，SIOS平均响应时间分别减少18.4%、25.8%、14.9%、14.5%和13.1%，而对于MLC固态盘，平均响应时间分别减少16.9%、24.4%、13.1%、13.0%和13.7%，结果表明，SIOS能有效减少I/O请求的平均响应时间，提高固态盘存储系统的I/O性能。

关键词: 固态盘, 请求调度, 非对称读/写, 请求大小, 响应时间

Abstract:

The response times are linear with the request sizes for flash-based Solid State Disk(SSD) with the same request type. Moreover, the read performance and write performance of flash-based SSD are asymmetric. Based on these characteristics, this paper proposes a Size-based I/O Scheduler(SIOS) for flash-based SSD to improve the I/O performance of SSD-based storage systems from the viewpoint of average response time. SIOS utilizes the asymmetric read and write performance characteristics of flash-based SSD and gives higher priority to the read requests. Moreover, by first processing the small requests in the I/O waiting queue, the average waiting times of the requests are reduced significantly. It implements SIOS in the Linux kernel and evaluates it with two kinds of SSD devices(SLC and MLC)driven by the five traces. Compared with the existing Linux disk I/O schedulers, evaluation results show that SIOS reduces average response times by 18.4%, 25.8%, 14.9%, 14.5% and 13.1% for SLC-based flash SSD, and reduces average response times by 16.9%, 24.4%, 13.1%, 13.0% and 13.7% for MLC-based flash SSD. Results show that compared with the state-of-the-arts, SIOS reduces the average response times significantly. Consequently, the I/O performance of the SSD-based storage systems is improved.

Key words: Solid State Disk(SSD), request scheduling, asymmetric read/write, request size, response time

中图分类号: 

• TP391