基于NAND闪存存储器的一种超级块闪存翻译层机制研究
|
摘要
NAND闪存的存储系统中,软件中间层闪存翻译层通常被用来隐藏写前必须先擦除的NAND闪存特性。介绍一种新的闪存翻译层机制,它会将相邻的逻辑块合并为一个块,称之为超级块,该超级块在闪存翻译层机制下是粗颗粒度映射,而其中的页是被细颗粒度映射到物理块的任意位置。为了减少额外的存储和闪存操作,细颗粒度映射信息被保存在NAND闪存的空闲区域。使用该混合映射技术不仅具有细颗粒度的灵活性而且只需要占用粗颗粒度的存储空间。实验结果研究表明,与其他闪存翻译层技术相比,该闪存翻译层机制减少了多达40%的垃圾回收操作。
In storage system of NAND flash memory,the flash translation layer,which is the intermediate layer of software,is usually used to hide the NAND flash memory property of erasing before writing. This paper introduces a new flash translation layer scheme,it will combine the adjacent logic blocks into one block,namely the superblock,the superblock is the coarse-grained mapping in flash translation layer scheme,and the pages in the superblock are mapped to arbitrary positions of physical block by the fine granularity. In order to reduce the extra memory and flesh memory operation,the fine-grained mapping information is stored in the free area of NAND flash memory. The use of the hybrid mapping technique has the flexibility of fine granularity,but only requires occupying the storage space of coarse granularity.Experimental results show that compared with other flash translation layer technologies,this scheme reduces the garbage collection operation up to 40%.
In storage system of NAND flash memory,the flash translation layer,which is the intermediate layer of software,is usually used to hide the NAND flash memory property of erasing before writing. This paper introduces a new flash translation layer scheme,it will combine the adjacent logic blocks into one block,namely the superblock,the superblock is the coarse-grained mapping in flash translation layer scheme,and the pages in the superblock are mapped to arbitrary positions of physical block by the fine granularity. In order to reduce the extra memory and flesh memory operation,the fine-grained mapping information is stored in the free area of NAND flash memory. The use of the hybrid mapping technique has the flexibility of fine granularity,but only requires occupying the storage space of coarse granularity.Experimental results show that compared with other flash translation layer technologies,this scheme reduces the garbage collection operation up to 40%.
引文
[1]Ban A.Flash file system:United States Patent,5404485[P].1995-4-4.
[2]Douglis F,Caceres R,Kaashoek F,et al.Tauber.Storage alternatives for mobilecomputers[C]//Proceedings of the First Symposium on Operating Systems Design and Implementation(OSDI-1),1994:25-37.
[3]邵亚刚,戴冠中,刘航.基于日志式混合映射的FTL算法设计与实现[J].计算机测量与控制,2009(7):1362-1364.
[4]牛伟,张延园.基于NAND Flash的文件系统设计与实现[J].计算机工程,2009(16):248-250.
[5]Harari E,Norman R D,Mehrota S.Flasheeprom system:United States Patent,5602987[P].February 1997.
[6]Kim H J,Lee S G.An effective flash memory manager for reli-able flash memory space management[J].IEICE Transactions on Information and System,2002,6:951-964.
[7]Inc M T.Small block vs.large block nand flashdevices[R].Technical report.Technical Note TN-29-07,2005.
[8]Kawaguchi A,Nishioka S,Motoda H.Aflash-memory based file system[C]//Proceedings of the USENIX Winter Technical Conference,1995:155-164.
[9]Kim J,Kim J M,Noh S,et al.Aspace-efficient flash translation layer for compactflashsystems[J].IEEE Transactions on Consumer Electronics,2002,48(2):366-375.
[10]岳丽华,刘沾沾,金培权,等.高效的闪存空间分配与回收方法[J].小型微型计算机系统,2010(5):819-825.
[2]Douglis F,Caceres R,Kaashoek F,et al.Tauber.Storage alternatives for mobilecomputers[C]//Proceedings of the First Symposium on Operating Systems Design and Implementation(OSDI-1),1994:25-37.
[3]邵亚刚,戴冠中,刘航.基于日志式混合映射的FTL算法设计与实现[J].计算机测量与控制,2009(7):1362-1364.
[4]牛伟,张延园.基于NAND Flash的文件系统设计与实现[J].计算机工程,2009(16):248-250.
[5]Harari E,Norman R D,Mehrota S.Flasheeprom system:United States Patent,5602987[P].February 1997.
[6]Kim H J,Lee S G.An effective flash memory manager for reli-able flash memory space management[J].IEICE Transactions on Information and System,2002,6:951-964.
[7]Inc M T.Small block vs.large block nand flashdevices[R].Technical report.Technical Note TN-29-07,2005.
[8]Kawaguchi A,Nishioka S,Motoda H.Aflash-memory based file system[C]//Proceedings of the USENIX Winter Technical Conference,1995:155-164.
[9]Kim J,Kim J M,Noh S,et al.Aspace-efficient flash translation layer for compactflashsystems[J].IEEE Transactions on Consumer Electronics,2002,48(2):366-375.
[10]岳丽华,刘沾沾,金培权,等.高效的闪存空间分配与回收方法[J].小型微型计算机系统,2010(5):819-825.