透過您的圖書館登入
IP:3.128.205.109
  • 學位論文

隨機寫入下的多通道映射快取最佳化方法

Optimizing mapping cache performance for multichannel SSDs under random-writing workloads

指導教授 : 張立平

摘要


固態硬碟擁有比傳統硬碟更快速的存取速度、較低的功率消耗與更好的抗震能力,隨著目前固態硬碟技術逐漸成熟而價格下跌,固態硬碟顯然已經成為儲存設備的主流。在固態硬碟的技術中,會使用映射快取的技術,透過將映射表以快取的形式,快取較常用的映射資訊,不常用的映射資訊存在快閃記憶體中,以減少映射表過大的問題,並且利用多通道架構,以多個獨立的通道服務要求,增加系統的效能。然而在隨機寫入的工作負荷下,映射快取與通道之間的相依性會造成嚴重的通道閒置問題,使得固態硬碟的效能提升受到限制。最佳化映射快取的相關文獻沒有考量到多通道架構下所產生的通道閒置問題,因此本篇論文提出最佳化映射快取演算法(Optimizing Mapping Cache, OMC),透過預先讀取與區分工作的優先權方式,消除映射快取在多通道架構下因為資料相依性而產生通道閒置而造成效能衰減的問題。

並列摘要


There are several advantages for solid state disks, compared to hard disk drives, such as faster access speed, lower power consumption, and higher shock resistance. Recently, technology of solid state disks gets mature and price becomes cheaper. Solid state disks become the mainstream of storage devices. In this technology, manufacturers use the mapping cache to deal with the problem of excessive mapping table, and increase their effectiveness using multi-channel architecture. In random workloads, causing idle channel because of the channel dependency. The performance boost is limited. In the related works, they didn’t consider the issues arising from the multi-channel architecture, so this paper proposes method called Optimizing Mapping Cache (OMC) using the concept of pre-reading and distinguishing priority of jobs to eliminate the channel idle time.

參考文獻


[1] Gupta A, Kim Y, Urgaonkar B. DFTL: a flash translation layer employing demand-based selective caching of page-level address mappings[M]. ACM, 2009.
[2] Li-Pin Chang and Tei-Wei Kuo. 2002. An Adaptive Striping Architecture for Flash Memory Storage Systems of Embedded Systems. In Proceedings of the Eighth IEEE Real-Time and Embedded Technology and Applications Symposium (RTAS'02) (RTAS '02). IEEE Computer Society, Washington, DC, USA, 187-.
[3] Qin Z, Wang Y, Liu D, et al. A two-level caching mechanism for demand-based page-level address mapping in NAND flash memory storage systems[C]//Real-Time and Embedded Technology and Applications Symposium (RTAS), 2011 17th IEEE. IEEE, 2011: 157-166.
[4] Jiang S, Zhang L, Yuan X H, et al. S-ftl: An efficient address translation for flash memory by exploiting spatial locality[C]//Mass Storage Systems and Technologies (MSST), 2011 IEEE 27th Symposium on. IEEE, 2011: 1-12.
[7] Park D, Debnath B, Du D. CFTL: A convertible flash translation layer adaptive to data access patterns[C]//ACM SIGMETRICS Performance Evaluation Review. ACM, 2010, 38(1): 365-366.

延伸閱讀