Chin-Hsien Wu
Abstract
When applications require high I/O performance, solid-state drives (SSDs) are often preferable because they perform better than traditional hard-disk drives (HDDs). Therefore, database system response time can be improved by moving frequently used data to SSDs. However, because capacity is both more limited and more expensive for SSDs than it is for HDDs, there is a need to determine which data are most appropriate to migrate from HDDs to SSDs. In this paper, we propose a data management method for databases using SSDs by considering an integrated mechanism and a migration rule for performing migrations between HDDs and SSDs. The proposed method can provide database systems with high I/O performance by properly moving high-priority data to SSDs with fast access capabilities while maintaining the low-priority data on HDDs, which have lower costs. The results of experiments show that the proposed method can achieve the goal while requiring only small amounts of SSD capacity.
- B-cache Repository. http://bcache.evilpiepirate.org/.Google Scholar
- FlashCache Repository. https://github.com/facebook/flashcache.Google Scholar
- Hammerdb is an open source database load testing and benchmarking tool.Google Scholar
- Tpc-c is an on-line transaction processing benchmark.Google Scholar
- Tpc-h is a decision support benchmark.Google Scholar
- Performance value of solid state drives using IBM i Repository. http://www-03.ibm.com/systems/resources/, May 2009.Google Scholar
- B. H. Bloom. Space/time trade-offs in hash coding with allowable errors. volume 13(7), pages 422--426, July 1970. Google ScholarDigital Library
- M. Canim, G. A. Mihaila, B. Bhattacharjee, K. A. Ross, and C. A. Lang. An object placement advisor for db2 using solid state storage. In Proceedings of the VLDB Endowment, volume 2, pages 1318--1329, Aug 2009. Google ScholarDigital Library
- M. Canim, G. A. Mihaila, B. Bhattacharjee, K. A. Ross, and C. A. Lang. Ssd bufferpool extensions for database systems. In Proceedings of the VLDB Endowment, volume 3, pages 1435--1446, September 2010. Google ScholarDigital Library
- J. Choi, B. Lee, D. Jung, and H. Y. Youn. An ssd-based accelerator using partitioned bloom filter for directory parsing. In IEEE International Conference on IT Convergence and Security, pages 1--5, August 2015.Google ScholarCross Ref
- J. Goldstein and P. ÃĚke Larson. Optimizing queries using materialized views: a practical, scalable solution. In ACM SIGMOD international conference on Management of data, pages 331--342, June 2001. Google ScholarDigital Library
- J. W. Hsieh, L. P. Chang, and T. W. Kuo. Efficient identification of hot data for flash memory storage systems. ACM Transactions on Storage, 2(1), 2006. Google ScholarDigital Library
- C.-K. Kang, Y.-J. Cai, C.-H. Wu, , and P.-C. Hsiu. A hybrid storage access framework for high-performance virtual machines. ACM Transactions on Embedded Computing Systems, 13(5s), 2014. Google ScholarDigital Library
- A. Katsifodimos, I. Manolescu, and V. Vassalos. Materialized view selection for xquery workloads. In ACM SIGMOD International Conference on Management of Data, pages 565--576, May 2012. Google ScholarDigital Library
- L. Lin, Y. Zhu, J. Yue, Z. Cai, and B. Segee. Hot random off-loading: A hybrid storage system with dynamic data migration. In 19th Annual IEEE International Symposium on Modelling, Analysis, and Simulation of Computer and Telecommunication Systems, pages 318--325, July 2011. Google ScholarDigital Library
- M. Mehta and D. J. DeWitt. Data placement in shared-nothing parallel database systems. volume 6, pages 53--72, February 1997. Google ScholarDigital Library
- Oracle. Take the guesswork out of database layout and i/o tuning with automatic storage management. In Oracle Technical White Paper, December 2005.Google Scholar
- J. Ou, J. Shu, Y. Lu, L. Yi, , and W. Wang. Edm: an endurance-aware data migration scheme for load balancing in ssd storage clusters. In IEEE International Parallel and Distributed Processing Symposium, pages 787--796, May 2014. Google ScholarDigital Library
- D. Park. Hot and cold data identification: Applications to storage devices and systems. In Ph.D. dissertation, UNIVERSITY OF MINNESOTA, 2012. Google ScholarDigital Library
- D. Park and D. H. Du. Hot data identification for flash-based storage systems using multiple bloom filters. In Mass Storage Systems and Technologies (MSST), pages 1--11, May 2011. Google ScholarDigital Library
- A. Sachedina, M. Huras, and A. Colangelo. Best practices database storage. In White paper, IBM DB2 for Linux, UNIX, and Windows, Oct 2008.Google Scholar
- J. Schindler, A. Ailamaki, and G. R. Ganger. Matching database access patterns to storage characteristics. In In FAST '02:Proceedings of the 1st USENIX Conference on Fileand Storage Technologies, page 22, 2002.Google Scholar
- H. Shi, R. V. Arumugam, C. H. Foh, and K. K. Khaing. Optimal disk storage allocation for multitier storage system. IEEE Transactions on magnetics, 49(6), June 2013.Google Scholar
- C.-H. Wu, P.-H. Wu, K.-L. Chen, W.-Y. Chang, and K.-C. Lai. A hotness filter of files for reliable non-volatile memory systems. IEEE Transactions on Dependable and Secure Computing, 12(4), July 2015.Google ScholarCross Ref
- C. Xu, W. Wang, D. Zhou, and T. Xie. An ssd-hdd integrated storage architecture for write-once-read-once applications on clusters. In IEEE International Conference on Cluster Computing, pages 74--77, September 2015. Google ScholarDigital Library
- C. Zhu, Q. Zhu, C. Zuzarte, and W. Ma. A materialized-view based technique to optimize progressive queries via dependency analysis. In Proceedings of the Conference of the Center for Advanced Studies on Collaborative Research, pages 60--73, November 2011. Google ScholarDigital Library
Index Terms
- A data management method for databases using hybrid storage systems
Recommendations
Sibyl: adaptive and extensible data placement in hybrid storage systems using online reinforcement learning
ISCA '22: Proceedings of the 49th Annual International Symposium on Computer ArchitectureHybrid storage systems (HSS) use multiple different storage devices to provide high and scalable storage capacity at high performance. Data placement across different devices is critical to maximize the benefits of such a hybrid system. Recent research ...
Hybrid storage management for database systems
The use of flash-based solid state drives (SSDs) in storage systems is growing. Adding SSDs to a storage system not only raises the question of how to manage the SSDs, but also raises the question of whether current buffer pool algorithms will still ...
A priority-based data placement method for databases using solid-state drives
RACS '18: Proceedings of the 2018 Conference on Research in Adaptive and Convergent SystemsWhen applications require high I/O performance, solid-state drives (SSDs) are often preferable because they perform better than traditional hard-disk drives (HDDs). Therefore, database system response time can be improved by moving frequently used data ...
Comments