Advertisement

SFCM: A SSD-Friendly Cache Management Policy for Hybrid Storage Systems

  • Jiangtao Wang
  • Wenyu Lai
  • Xiaofeng Meng
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7901)

Abstract

With the growing popularity of flash memory, solid state drivers (SSD) based on flash memory have been widely used in various kinds of applications. Compared with conventional magnetic disk, SSD can provide higher access bandwidth and lower access latency. However, it will not completely replace the disk as the secondary storage in the short run due to its inherent properties such as asymmetric read/write, high price of per gigabyte etc. Integrating SSD and magnetic disk together can make full use of different performance advantages, so as to obtain good high performance and low cost. This paper proposes SSD-friendly cache management scheme (SFCM) that use SSD as a cache layer between main memory and magnetic disk. For the pages evicted by buffer manager, SFCM conditionally caches them in SSD according to the state of the page and the different replacement cost. Due to the quick access performance of SSD, SFCM can improve the performance of cache management when the destination page resides in SSD. In view of the poor random write performance for SSD, SFCM adopts flash-aware algorithms and data structures to manage the data stored in SSD, which takes care of write patterns and request types’ impacts on system performance. Furthermore, the study proposes a flash-friendly buffer replacement policy which considers the asymmetric I/O cost and the workload access features. We implement the scheme on the multi-level cache storage system based on a simulation platform and evaluate the performance. The experimental results show that SFCM can significantly reduce system response time and disk I/O cost. Adding a flash-based SSD as cache extension can make up the performance gap between memory and disk.

Keywords

Solid State Driver Cache Flash-Aware Replacement Cost 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Canim, M., Bhattacharjee, B., Mihaila, G.A., Lang, C.A., Ross, K.A.: An Object Placement Advisor for DB2 Using Solid State Storage. Proceedings of the Very Large Data Base (VLDB) Endowment 2(2), 1318–1329 (2009)Google Scholar
  2. 2.
    D. White Paper: MySpace Uses Fusion Powered I/O to Drive Greener and Better Data Centers.(EB/OL), http://www.fusionio.com/case-studies/myspace-case-study.pdf
  3. 3.
    Lee, S., Moon, B., Park, C., Hwang, J., Kim, K.: Accelerating In-Page Logging with Non-Volatile Memory. Data Engineering 33(4), 41–47 (2010)Google Scholar
  4. 4.
    Debnath, B., Sengupta, S., Li, J.: Flashstore: high throught persistent key-value store. Proceedings of the Very Large Data Base (VLDB) Endowment 3(2), 1414–1425 (2010)Google Scholar
  5. 5.
    Koltsidas, I., Viglas, S.D.: Designing a Flash-Aware Two-Level Cache. In: Eder, J., Bielikova, M., Tjoa, A.M. (eds.) ADBIS 2011. LNCS, vol. 6909, pp. 153–169. Springer, Heidelberg (2011)CrossRefGoogle Scholar
  6. 6.
    Mustafa, C., George, M., Bishwaranjan, B., Kenneth, R., Christian, L.: SSD Bufferpool Extensions for Database Systems. In: VLDB, pp. 1435–1446 (2010)Google Scholar
  7. 7.
    Kang, W.H., Lee, S.W., Moon, B.: Flash-based Extended Cache for Higher Throughput and Faster Recovery. Proceedings of the Very Large Data Base (VLDB) Endowment 5(11), 1615–1626 (2012)Google Scholar
  8. 8.
    Chen, F., Koufaty, D.A., Zhang, X.: Understanding intrinsic characteristics and system implications of flash memory based solid state drives. In: SIGMETRICS, pp. 181–192 (2009)Google Scholar
  9. 9.
    Bouganim, L., Jónsson, B.T., Bonnet, P.: uFLIP:Understanding flash io patterns. In: CIDR (2009)Google Scholar
  10. 10.
    Muntz, D., Honeyman, P.: Multi-Level Caching in Distributed File Systems-or-Your Cache Ain’t Nuthin’ but Trash. In: USENIX, pp. 305–314 (1992)Google Scholar
  11. 11.
    Willick, D.L., Eager, D.L., Bunt, R.B.: Disk Cache Replacement Policies for Network Fileservers. In: ICDCS, pp. 2–11 (1993)Google Scholar
  12. 12.
    Ganger, G., Worthington, B., et al.: The DiskSim simulation environment(v4.o) (E13/oL] (September 2009), http://www.pdl.ainu.ed/DiskSim
  13. 13.
    Agrawal, N., Prabhakaran, V., Wobber, T., et al.: Design tradeoffs for SSD performance. In: USENLX, pp. 57–70 (2008)Google Scholar
  14. 14.
    TPC Benchmark C: Standard Specification, http://www.tpc.org/tpcc/spec/tpcccurrent.pdf

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Jiangtao Wang
    • 1
  • Wenyu Lai
    • 1
  • Xiaofeng Meng
    • 1
  1. 1.Renmin University of ChinaBeijingChina

Personalised recommendations