Research On Channel-RAID For An SSD

NAND flash memory based solid state drives (hereafter, SSDs) possess some attractive properties such as high performance, low energy consumption, high shock resistance, and small physical size compared to traditional hard disk drives (HDDs). Manufacturers are continuously pushing NAND flash memory into smaller geometries and enforce each cell to store multiple bits in order to largely reduce its cost. Unfortunately, these scaling down techniques inherently degrade the endurance and reliability of flash memory. As a result, permanent errors such as block or die failures could occur with a higher possibility. While most transient errors like programming errors and read errors can be fixed by an ECC (error correction code) scheme, rectifying permanent errors requires a data redundancy mechanism like RAID (Redundant Array of Independent Disks) in a single SSD where multiple channels work in parallel.This paper makes research on RAID mechanism on the channel level of a single SSD. Firstly, we analyze and compare various RAID mechanisms applied on SSDs. Secondly, we first implement several common RAID structures in the channel level of a single SSD to understand their impact on an SSD’s performance. Next, to enhance the reliability of a SSD while maintaining its performance, we propose a new data redundancy architecture called CR5M (Channel-RAID5 with Mirroring), which can be applied to one SSD for mission-critical applications. CR5M utilizes hidden mirror chips to accelerate the performance of small writes. Finally, we extend the SSDSim which is an event-driven, modularly structured, and highly accurate simulator for SSD to support a handful of channel-level RAID architectures such as channel-RAID 1 (CR1), channel-RAID4 (CR4), channel-RAID5 (CR5), and channel-level RAID5 with mirroring (CR5M). We conduct extensive simulations using real-world traces and synthetic benchmarks on a extended simulator to evaluate CRs. Experimental results demonstrate that compared with CR5 (Channel-RAID5) CR5M decreases mean response time by up to 25.8%. Besides, it reduces the average writes per channel by up to 23.6%.