Wei Yi, Z. Sun, Hui Xu, Jietao Diao, Nan Li, Mingqian Wang
{"title":"双RAID:一种高可靠性的全闪存阵列方案","authors":"Wei Yi, Z. Sun, Hui Xu, Jietao Diao, Nan Li, Mingqian Wang","doi":"10.1109/JCSSE.2014.6841870","DOIUrl":null,"url":null,"abstract":"Flash-based solid state disks (SSDs) are replacing hard disk drives owing to its high performance and low energy consumption. For both high reliability and large capacity, redundant array of independent disks (RAID) is commonly used in storage system. In all flash arrays, the HDDs in RAID are substituted by SSDs and it gains improvements in bandwidth and IOPS (Input/output Operations per Second) by leaps and bounds owing to SSDs high performance. However, the error mode of SSD is different from HDD. The uncorrectable page errors in SSD are a great threat especially when a disk fails. For high reliability, we proposed a Dual-RAID scheme to protect data at disk level and block level respectively. To correct the disk failure, the proposed scheme adopts RAID-5 between disks. For recovering uncorrectable page errors, a redundant array of independent NANDs (RAIN) scheme is applied to SSD by exploiting its multi-channel structure. Although the redundancy overhead of the proposed scheme is about the same as RAID-6, the dual-RAID scheme achieves better performance and higher reliability. The experiment shows that the proposed scheme improves performance 35% on totally random trace in comparison to RAID 6 scheme.","PeriodicalId":331610,"journal":{"name":"2014 11th International Joint Conference on Computer Science and Software Engineering (JCSSE)","volume":"388 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2014-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"Dual RAID: A scheme for high reliable all flash array\",\"authors\":\"Wei Yi, Z. Sun, Hui Xu, Jietao Diao, Nan Li, Mingqian Wang\",\"doi\":\"10.1109/JCSSE.2014.6841870\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Flash-based solid state disks (SSDs) are replacing hard disk drives owing to its high performance and low energy consumption. For both high reliability and large capacity, redundant array of independent disks (RAID) is commonly used in storage system. In all flash arrays, the HDDs in RAID are substituted by SSDs and it gains improvements in bandwidth and IOPS (Input/output Operations per Second) by leaps and bounds owing to SSDs high performance. However, the error mode of SSD is different from HDD. The uncorrectable page errors in SSD are a great threat especially when a disk fails. For high reliability, we proposed a Dual-RAID scheme to protect data at disk level and block level respectively. To correct the disk failure, the proposed scheme adopts RAID-5 between disks. For recovering uncorrectable page errors, a redundant array of independent NANDs (RAIN) scheme is applied to SSD by exploiting its multi-channel structure. Although the redundancy overhead of the proposed scheme is about the same as RAID-6, the dual-RAID scheme achieves better performance and higher reliability. The experiment shows that the proposed scheme improves performance 35% on totally random trace in comparison to RAID 6 scheme.\",\"PeriodicalId\":331610,\"journal\":{\"name\":\"2014 11th International Joint Conference on Computer Science and Software Engineering (JCSSE)\",\"volume\":\"388 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2014-05-14\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2014 11th International Joint Conference on Computer Science and Software Engineering (JCSSE)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/JCSSE.2014.6841870\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2014 11th International Joint Conference on Computer Science and Software Engineering (JCSSE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/JCSSE.2014.6841870","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
摘要
基于闪存的固态硬盘(ssd)以其高性能、低能耗的优点正在逐渐取代硬盘。RAID (redundant array of independent disks)是存储系统中常用的一种存储方式,具有高可靠性和大容量的双重要求。在所有的闪存阵列中,RAID中的hdd都被ssd取代,由于ssd的高性能,带宽和IOPS (Input/output Operations per Second)都有了跳跃式的提升。但SSD和HDD的错误模式不同。在SSD中,不可纠正的页面错误是一个很大的威胁,特别是当磁盘出现故障时。为了提高可靠性,我们提出了双raid方案,分别在磁盘级和块级保护数据。为了纠正磁盘故障,本方案在磁盘之间采用RAID-5。为了恢复不可纠正的页面错误,利用固态硬盘的多通道结构,将独立NANDs冗余阵列(RAIN)方案应用于固态硬盘。虽然该方案的冗余开销与RAID-6大致相同,但双raid方案具有更好的性能和更高的可靠性。实验表明,与RAID 6方案相比,该方案在完全随机跟踪上的性能提高了35%。
Dual RAID: A scheme for high reliable all flash array
Flash-based solid state disks (SSDs) are replacing hard disk drives owing to its high performance and low energy consumption. For both high reliability and large capacity, redundant array of independent disks (RAID) is commonly used in storage system. In all flash arrays, the HDDs in RAID are substituted by SSDs and it gains improvements in bandwidth and IOPS (Input/output Operations per Second) by leaps and bounds owing to SSDs high performance. However, the error mode of SSD is different from HDD. The uncorrectable page errors in SSD are a great threat especially when a disk fails. For high reliability, we proposed a Dual-RAID scheme to protect data at disk level and block level respectively. To correct the disk failure, the proposed scheme adopts RAID-5 between disks. For recovering uncorrectable page errors, a redundant array of independent NANDs (RAIN) scheme is applied to SSD by exploiting its multi-channel structure. Although the redundancy overhead of the proposed scheme is about the same as RAID-6, the dual-RAID scheme achieves better performance and higher reliability. The experiment shows that the proposed scheme improves performance 35% on totally random trace in comparison to RAID 6 scheme.