共享儲存系統由於有著更好的空間利用度及較低的硬體維護成本等優點，已經成為一種普遍使用的儲存系統方案。近來由於固態硬碟有優於傳統硬碟的一些特性使得固態硬碟更為廣泛地被使用在共享儲存系統中。也因為固態硬碟的生產成本逐年下降，使得共享儲存系統中能夠使用多顆的固態硬碟。
在使用共享儲存裝置時，效能隔離是一個重要的議題，藉由滿足每位使用者提出的服務目標(Service Level Objectives, SLO)是達成效能隔離的方法，但是使用者之間會有干擾現象的存在導致效能隔離無法達成。而VSSD的提出，能夠有效的解決干擾達成效能隔離，但將VSSD直接應用到使用多顆固態硬碟的共享儲存系統時，發現了資源分配與負載不相符導致效能不佳的現象。本篇論文提出了一個動態調整資源分配的方法，稱為Libra。Libra基於VSSD的架構擴展而來，能夠保有效能隔離的情況下提升系統平均效能。
實驗結果顯示Libra相較於VSSD可以提升系統平均效能達 15 %，另外也能夠平衡共享儲存系統中多顆固態硬碟的壽命消耗。

Shared storage system has become a popular storage solution because the advantages such as better utilization, lower hardware and maintenance costs. Recently, solid-state drive (SSD) has factors superior to hard disk drive (HDD) and has been increasingly used in shared storage system. Because the production costs of SSD is decreasing in numbers every year, shared storage system contains multi-SSD.
Performance isolation is critical in shared storage system. By satisfying the service level object (SLO) required by users is a method to achieve performance isolation. But there are interferences between users results. The existence of interference results in that performance isolation cannot be achieve. A framework called VSSD is proposed to solve these interferences to keep performance isolation in a shared storage. However, VSSD cannot be applied to a shared storage system with multi-SSD directly because the poor performance caused by inefficient resource distribution. In order to achieve efficient resource utilization, a framework called Libra extended from VSSD is proposed.
Experiment results show that Libra can improve by up to 36% (15% on average) relative to original VSSD and balance the lifetime consume between the SSDs.

[1] N. Agrawal, V. Prabhakaran, T. Wobber, J. D. Davis, M. S. Manasse, and R. Panigrahy, "Design Tradeoffs for SSD Performance," in USENIX Annual Technical Conference, 2008, vol. 8, pp. 57-70.
[2] L.-P. Chang, T.-W. Kuo, and S.-W. Lo, "Real-time garbage collection for flash-memory storage systems of real-time embedded systems," ACM Transactions on Embedded Computing Systems (TECS), vol. 3, no. 4, pp. 837-863, 2004.
[3] Q. Zhang, X. Li, L. Wang, T. Zhang, Y. Wang, and Z. Shao, "Optimizing deterministic garbage collection in NAND flash storage systems," in Real-Time and Embedded Technology and Applications Symposium (RTAS), 2015 IEEE, 2015, pp. 14-23: IEEE.
[4] D.-W. Chang, H.-H. Chen, D.-J. Yang, and H.-P. Chang, "BLAS: Block-level adaptive striping for solid-state drives," ACM Transactions on Design Automation of Electronic Systems (TODAES), vol. 19, no. 2, p. 21, 2014.
[5] P. Desnoyers, "Analytic modeling of SSD write performance," in Proceedings of the 5th Annual International Systems and Storage Conference, 2012, p. 12: ACM.
[6] X.-Y. Hu, E. Eleftheriou, R. Haas, I. Iliadis, and R. Pletka, "Write amplification analysis in flash-based solid state drives," in Proceedings of SYSTOR 2009: The Israeli Experimental Systems Conference, 2009, p. 10: ACM.
[7] W.-H. Lin and L.-P. Chang, "Dual greedy: Adaptive garbage collection for page-mapping solid-state disks," in Design, Automation & Test in Europe Conference & Exhibition (DATE), 2012, 2012, pp. 117-122: IEEE.
[8] J.-W. Hsieh, L.-P. Chang, and T.-W. Kuo, "Efficient on-line identification of hot data for flash-memory management," in Proceedings of the 2005 ACM symposium on Applied computing, 2005, pp. 838-842: ACM.
[9] D. Park and D. H. Du, "Hot and cold data identification for flash memory using multiple bloom filters," in USENIX Conference on File and Storage Technologies (FAST), 2011.
[10] D.-W. Chang, H.-H. Chen, and W.-J. Su, "VSSD: Performance Isolation in a Solid-State Drive," ACM Transactions on Design Automation of Electronic Systems (TODAES), vol. 20, no. 4, p. 51, 2015.
[11] L.-P. Chang and T.-W. Kuo, "An adaptive striping architecture for flash memory storage systems of embedded systems," in Real-Time and Embedded Technology and Applications Symposium, 2002. Proceedings. Eighth IEEE, 2002, pp. 187-196: IEEE.
[12] S. Lee, D. Shin, Y.-J. Kim, and J. Kim, "LAST: locality-aware sector translation for NAND flash memory-based storage systems," ACM SIGOPS Operating Systems Review, vol. 42, no. 6, pp. 36-42, 2008.
[13] 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 Parallel and Distributed Processing Symposium, 2014 IEEE 28th International, 2014, pp. 787-796: IEEE.
[14] J. Tai, D. Liu, Z. Yang, X. Zhu, J. Lo, and N. Mi, "Improving flash resource utilization at minimal management cost in virtualized flash-based storage systems," IEEE Transactions on Cloud Computing, vol. 5, no. 3, pp. 537-549, 2017.
[15] Y. Du, Y. Zhang, and N. Xiao, "R-dedup: content aware redundancy management for ssd-based raid systems," in Parallel Processing (ICPP), 2014 43rd International Conference on, 2014, pp. 111-120: IEEE.
[16] Z. Yao, I. Papapanagiotou, and R. Griffith, "Serifos: Workload consolidation and load balancing for SSD based cloud storage systems," arXiv preprint arXiv:1512.06432, 2015.
[17] J. S. J. S. Bucy, S. W. Schlosser, and G. R. Ganger. (2008). The Disksim simulation environment version 4.0. .
[18] Microsoft. (2009). SSD Extension for DiskSim Simulation Environment. Available: http://research.microsoft.com/en-us/downloads/b41019e2-1d2b-44d8-b512-ba35ab814cd4/
[19] IOTTA, "Storage Networking Industry Association," 2015. Available: http://iotta.snia.org/