Pinpointing and scheduling access conflicts to improve internal resource utilization in solid-state drives

Abstract

Modern solid-state drives (SSDs) are integrating more internal resources to achieve higher capacity. Parallelizing accesses across internal resources can potentially enhance the performance of SSDs. However, exploiting parallelism inside SSDs is challenging owing to real-time access conflicts. In this paper, we propose a highly parallelizable I/O scheduler (PIOS) to improve internal resource utilization in SSDs from the perspective of I/O scheduling. Specifically, we first pinpoint the conflicting flash requests with precision during the address translation in the Flash Translation Layer (FTL). Then, we introduce conflict eliminated requests (CERs) to reorganize the I/O requests in the device-level queue by dispatching conflicting flash requests to different CERs. Owing to the significant performance discrepancy between flash read and write operations, PIOS employs differentiated scheduling schemes for read and write CER queues to always allocate internal resources to the conflicting CERs that are more valuable. The small dominant size prioritized scheduling policy for the write queue significantly decreases the average write latency. The high parallelism density prioritized scheduling policy for the read queue better utilizes resources by exploiting internal parallelism aggressively. Our evaluation results show that the parallelizable I/O scheduler (PIOS) can accomplish better SSD performance than existing I/O schedulers implemented in both SSD devices and operating systems.

Keywords

solid-state drive / access conflict / I/O scheduler / internal resource utilization / PIOS

References

[1] Agrawal N, Prabhakaran V, Wobber T, Davis J D, Manasse M S, Panigrahy R. Design tradeoffs for SSD performance. In: Proceedings of USENIX Annual Technical Conference. 2008, 57–70

[2] Chen F, Koufaty D A, Zhang X. Hystor: making the best use of solid state drives in high performance storage systems. In: Proceedings of the International Conference on Supercomputing. 2011, 22–32

[3] Saxena M, Swift M M, Zhang Y. FlashTier: a lightweight, consistent and durable storage cache. In: Proceedings of the 7th ACM European Conference on Computer Systems. 2012, 267–280

[4] Caulfield A M, Grupp L M, Swanson S. Gordon: using flash memory to build fast, power-efficient clusters for data-intensive applications. ACM SIGPLAN Notices, 2009, 44(3): 217–228

[5] Kim H J, Lee Y S, Kim J S. NVMeDirect: a user-space I/O framework for application-specific optimization on NVMe SSDs. In: Proceedings of the 8th USENIX Workshop on Hot Topics in Storage and File Systems. 2016

[6] Xu W, Lu Y, Li Q, Zhou E, Song Z, Dong Y, Zhang W, Wei D, Zhang X, Chen H, Xing J, Yuan Y. Hybrid hierarchy storage system in milkyway- 2 supercomputer. Frontiers of Computer Science, 2014, 8(3): 367–377

[7] Liao X, Xiao L, Yang C, Lu Y. Milkyway-2 supercomputer: system and application. Frontiers of Computer Science, 2014, 8(3): 345–356

[8] Wires J, Ingram S, Drudi Z, Harvey N J, Warfield A. Characterizing storage workloads with counter stacks. In: Proceedings of the 11th USENIX Symposium on Operating Systems Design and Implementation. 2014, 335–349

[9] Gupta A, Kim Y, Urgaonkar B. DFTL: a flash translation layer employing demand-based selective caching of page-level address mappings. In: Proceedings of the 14th International Conference on Architecture Support for Programming Languages and Operating Systems. 2009, 229–240

[10] Xie X, Li Q, Wei D, Song Z, Xiao L. ECAM: an efficient cache management strategy for address mappings in flash translation layer. In: Proceedings of the International Workshop on Advanced Parallel Processing Technologies. 2013, 146–159

[11] Park D, Debnath B, Du D. CFTL: an adaptive hybrid flash translation layer with efficient caching strategies. IEEE Transactions on Computers, 2011, 1–15

[12] Yang M C, Chang Y M, Tsao C W, Huang P C, Chang Y H, Kuo T W. Garbage collection and wear leveling for flash memory: past and future. In: Proceedings of the International Conference on Smart Computing (SMARTCOMP). 2014, 66–73

[13] Yang M C, Chang Y H, Tsao C W, Huang P C. New era: new efficient reliability-aware wear leveling for endurance enhancement of flash storage devices. In: Proceedings of the 50th Annual Design Automation Conference. 2013

[14] Hu Y, Jiang H, Feng D, Tian L, Luo H, Ren C. Exploring and exploiting the multilevel parallelism inside ssds for improved performance and endurance. IEEE Transactions on Computers, 2013, 62(6): 1141–1155

[15] Jung M, Kandemir M T. Sprinkler: maximizing resource utilization in many-chip solid state disks. In: Proceedings of the 20th IEEE International Symposium on High Performance Computer Architecture. 2014, 524–535

[16] Chen F, Koufaty D A, Zhang X. Understanding intrinsic characteristics and system implications of flash memory based solid state drives. ACM SIGMETRICS Performance Evaluation Review, 2009, 37(1): 181–192

[17] Jung M. Exploring parallel data access methods in emerging nonvolatile memory systems. IEEE Transactions on Parallel and Distributed Systems, 2017, 28(3): 746–759

[18] Chen F, Lee R, Zhang X. Essential roles of exploiting internal parallelism of flash memory based solid state drives in high-speed data processing. In: Proceedings of the 17th IEEE International Symposium on High Performance Computer Architecture. 2011, 266–277

[19] Jung M, Wilson III E H, Kandemir M. Physically addressed queueing (PAQ): improving parallelism in solid state disks. ACM SIGARCH Computer Architecture News, 2012, 40(3): 404–415

[20] Hu Y, Jiang H, Feng D, Tian L, Luo H, Zhang S. Performance impact and interplay of SSD parallelism through advanced commands, allocation strategy and data granularity. In: Proceedings of the International Conference on Supercomputing. 2011, 96–107

[21] Xie X, Wei D, Li Q, Song Z, Xiao L. CER-IOS: internal resource utilization optimized I/O scheduling for solid state drives. In: Proceedings of the 21st IEEE International Conference on Parallel and Distributed Systems. 2015, 336–343

[22] Gao C, Shi L, Zhao M, Xue C J, Wu K, Sha E H. Exploiting parallelism in I/O scheduling for access conflict minimization in flash-based solid state drives. In: Proceedings of the 30th Symposium on Mass Storage Systems and Technologies. 2014, 1–11

[23] Nam E H, Kim B S J, Eom H, Min S L. Ozone (O3): an out-of-order flash memory controller architecture. IEEE Transactions on Computers, 2011, 60(5): 653–666

[24] Tanenbaum A S. Modern Operating Systems. New Jersey: Prentice Hall, 2009

[25] Yu Y J, Shin D I, Eom H, Yeom H Y. NCQ vs. I/O scheduler: preventing unexpected misbehaviors. ACM Transactions on Storage, 2010, 6(1): 2

[26] Park S, Shen K. FIOS: a fair, efficient flash I/O scheduler. In: Proceedings of the 10th USENIX Conference on File and Storage Technologies. 2012

[27] Shen K, Park S. Flashfq: a fair queueing I/O scheduler for flash-based ssds. In: Proceedings of USENIX Annual Technical Conference. 2013, 67–78

[28] Guo J, Hu Y, Mao B, Wu S. Parallelism and garbage collection aware I/O scheduler with improved ssd performance. In: Proceedings of IEEE International Conference on Parallel and Distributed Processing Symposium. 2017, 1184–1193

[29] Narayanan D, Donnelly A, Rowstron A. Write off-loading: practical power management for enterprise storage. ACM Transactions on Storage, 2008, 4(3): 10

[30] Kim J, Oh Y, Kim E, Choi J, Lee D, Noh S H. Disk schedulers for solid state drivers. In: Proceedings of the 7th ACM International Conference on Embedded Software. 2009, 295–304

[31] Mao B, Wu S. Exploiting request characteristics and internal parallelism to improve ssd performance. In: Proceedings of the 33rd IEEE International Conference on Computer Design. 2015, 447–450

RIGHTS & PERMISSIONS

2018 Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature