Loading...

Performance Enhancement of Enterprise Storage Systems Using a Markov-Based Prefetching Method

Sereshki, Sina | 2012

533 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 43724 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Asadi, Hossein
  7. Abstract:
  8. With increasing rate of digital information in the world, design, configuration, and networking of enterprise storage systems has become an essential part in designing data centers. The performance of data storage systems in serving incoming requests is one of the major parameters of such systems. A major metric to measure performance is response time. This parameter is, in particular, crucial in enterprise applications such as financial, credit, multimedia, and real-time applications. A common approach to enhance the performance of enterprise storage systems is improving the hit ratio of the system global memory using prefetching technique. Using prefetching technique, a data block is transferred from Hard Disk Drives (HDDs) to the global memory before it is requested by the servers.In this thesis, we present a Markov based prefetching method. Using the proposed method, the probability of possible requesting data blocks is calculated and those with the most probability are prefetched to the global memory. To this end, the pattern of previous requests is stored using the proposed method and upon detecting similar pattern, target data blocks are prefetched. In particular, we employ Variable-Length Markov Chain (VLMC) to perform prefetching with various depth of Markov tree. To evaluate the efficiency of the proposed method, we implement VLMC using the DiskSim simulator. We also extract the hit ratio of the global memory using standard I/O workloads and compare it with previous prefetching methods. The results demonstrate that the proposed prefetching technique improves the hit ratio by 3.8 and 1.3 times as compared to the non-prefetching and table based prefetching methods, respectively.

  9. Keywords:
  10. Increasing Efficiency ; Data Center ; Data Prefetching ; Data Storage

 Digital Object List

 Bookmark

No TOC