Abstract
DM-SMR (device-managed shingled magnetic recording) disks allocate a portion of disk space as the persistent cache (PC) to address the issue of overlapping tracks during data updates. When the PC space becomes insufficient, a space cleaning is triggered to reclaim its invalid space. However, the space cleaning is time-consuming and contributes to the long-tail latency of DM-SMR disks. In the article, we will propose a space-grained cleaning method that leverages various idle periods to effectively reduce the long-tail latency of DM-SMR disks. The objective is to perform a proper space-grained cleaning for a suitable space region at an appropriate time period, thereby preventing delays in subsequent I/O requests and reducing the long-tail latency associated with DM-SMR disks. The experimental results demonstrate a substantial reduction in the long-tail latency of DM-SMR disks through the proposed method.
- [1] . 2015. Skylight-a window on shingled disk operation. ACM Transactions on Storage 11, 4(2015), 1–28.
DOI: Google ScholarDigital Library - [2] . 2010. Indirection systems for shingled-recording disk drives. In Proceedings of the 2010 IEEE 26th Symposium on Mass Storage Systems and Technologies. 1–14.
DOI: Google ScholarDigital Library - [3] . 2023. Short-term and long-term idle time detectors for reducing long-tail latency in solid-state drives. In Proceedings of the 2023 6th International Symposium on Computer, Consumer and Control. 143–146.
DOI: Google ScholarCross Ref - [4] . 2022. KVSTL: An application support to LSM-tree based key-value store via shingled translation layer data management. IEEE Transactions on Computers 71, 7 (2022), 1598–1611.
DOI: Google ScholarCross Ref - [5] . 2020. DSTL: A demand-based shingled translation layer for enabling adaptive address mapping on SMR drives. ACM Transactions on Embedded Computing Systems 19, 4(2020), 21 pages.
DOI: Google ScholarDigital Library - [6] . Seagate Datasheet. (n.d.). Retrieved December 18, 2023 from https://www.seagate.com/content/dam/seagate/migrated-assets/www-content/product-content/hdd-fam/seagate-archive-hdd/en-us/docs/archive-hdd-ds1834-5c-1508us.pdfGoogle Scholar
- [7] . 2009. Shingled recording for 2–3 Tbit/in\(^2\). IEEE Transactions on Magnetics 45, 10 (2009), 3823–3829.
DOI: Google ScholarCross Ref - [8] . 2020. uCache: A mutable cache for SMR translation layer. In Proceedings of the 2020 28th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems. 1–8.
DOI: Google ScholarCross Ref - [9] . 2017. SMaRT: An approach to shingled magnetic recording translation. In Proceedings of the 15th USENIX Conference on File and Storage Technologies. USENIX Association, Santa Clara, CA, 121–134. Retrieved from https://www.usenix.org/conference/fast17/technical-sessions/presentation/heGoogle Scholar
- [10] . 2010. Finite-element and micromagnetic modeling of write heads for shingled recording. IEEE Transactions on Magnetics 46, 3 (2010), 715–721.
DOI: Google ScholarCross Ref - [11] . 2017. Understanding storage traffic characteristics on enterprise virtual desktop infrastructure. In Proceedings of the 10th ACM International Systems and Storage Conference. 1–11.Google ScholarDigital Library
- [12] . 2023. HSMR-RAID: Enabling a low overhead RAID-5 system over a host-managed shingled magnetic recording disk array. In Proceedings of the 38th ACM/SIGAPP Symposium on Applied Computing.Association for Computing Machinery, New York, NY, USA, 294–296.
DOI: Google ScholarDigital Library - [13] . 2020. KFR: Optimal cache management with k-framed reclamation for drive-managed SMR disks. In Proceedings of the 2020 57th ACM/IEEE Design Automation Conference. 1–6.
DOI: Google ScholarCross Ref - [14] . 2016. HMSS: A high performance host-managed shingled storage system based on awareness of SMR on block layer. In Proceedings of the 2016 IEEE 18th International Conference on High Performance Computing and Communications; IEEE 14th International Conference on Smart City; IEEE 2nd International Conference on Data Science and Systems. 570–577.
DOI: Google ScholarCross Ref - [15] . 2008. Write off-loading: Practical power management for enterprise storage. ACM Transactions on Storage 4, 3 (2008), 1–23.Google ScholarDigital Library
- [16] . 2021. Reinforcement learning-assisted cache cleaning to mitigate long-tail latency in DM-SMR. In Proceedings of the 2021 58th ACM/IEEE Design Automation Conference. 103–108.
DOI: Google ScholarDigital Library - [17] . 2015. SMRDB: Key-value data store for shingled magnetic recording disks. In SYSTOR ’15: Proceedings of the 8th ACM International Systems and Storage Conference.Association for Computing Machinery, New York, NY, USA, 11 pages.
DOI: Google ScholarDigital Library - [18] . 2022. Large-Capacity, High-Performance 3.5-inch HDDs for Surveillance Camera Systems Applying SMR Technologies. (2022). Retrieved December 18, 2023 from https://toshiba.semicon-storage.com/content/dam/toshiba-ss-v3/master/en/company/technical-review/pdf/3_5-inch-hdd-smr_202303_en.pdfGoogle Scholar
- [19] . 2017. Virtual guard: A track-based translation layer for shingled disks. In Proceedings of the 9th USENIX Workshop on Hot Topics in Storage and File Systems. USENIX Association, Santa Clara, CA. Retrieved from https://www.usenix.org/conference/hotstorage17/program/presentation/shafaeiGoogle Scholar
- [20] . 2017. Modeling drive-managed SMR performance. ACM Transactions on Storage 13, 4(2017), 1–22.
DOI: Google ScholarDigital Library - [21] . 2018. Reinforcement Learning: An Introduction. MIT press, Cambridge, MA, USA. Retrieved from https://mitpress.mit.edu/books/reinforcement-learning-second-editionGoogle ScholarDigital Library
- [22] . 2016. Evaluating host aware SMR drives. In Proceedings of the 8th USENIX Workshop on Hot Topics in Storage and File Systems. USENIX Association, Denver, CO. Retrieved from https://www.usenix.org/conference/hotstorage16/workshop-program/presentation/wuGoogle ScholarDigital Library
- [23] . 2019. An active method to mitigate the long latencies for host-aware shingle magnetic recording drives. In Proceedings of the 2019 IEEE 25th International Conference on Parallel and Distributed Systems. 17–26.
DOI: Google ScholarCross Ref - [24] . 2017. Virtual persistent cache: Remedy the long latency behavior of host-aware shingled magnetic recording drives. In Proceedings of the 2017 IEEE/ACM International Conference on Computer-Aided Design. 17–24.
DOI: Google ScholarDigital Library - [25] . 2019. SEALDB: An efficient LSM-tree based KV store on SMR drives with sets and dynamic bands. IEEE Transactions on Parallel and Distributed Systems 30, 11 (2019), 2595–2607.
DOI: Google ScholarDigital Library - [26] . 2019. GearDB: A GC-free key-value store on HM-SMR drives with gear compaction. In Proceedings of the 17th USENIX Conference on File and Storage Technologies. USENIX Association, Boston, MA, 159–171. Retrieved from https://www.usenix.org/conference/fast19/presentation/yaoGoogle Scholar
- [27] . 2020. Idler : I/O workload controlling for better responsiveness on host-aware shingled magnetic recording drives. IEEE Transactions on Computers 69, 6 (2020), 777–788.
DOI: Google ScholarCross Ref - [28] . 2023. SMRSTORE: A storage engine for cloud object storage on HM-SMR drives. In Proceedings of the 21st USENIX Conference on File and Storage Technologies. USENIX Association, Santa Clara, CA, 395–408. Retrieved from https://www.usenix.org/conference/fast23/presentation/zhouGoogle Scholar
Index Terms
- A Space-Grained Cleaning Method to Reduce Long-Tail Latency of DM-SMR Disks
Recommendations
SMR Disks for Mass Storage Systems
MASCOTS '15: Proceedings of the 2015 IEEE 23rd International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication SystemsDisk drives have seen a dramatic increase in storage density over the last five decades, but to continue the growth seems difficult because of physical limitations. One promising approach to overcome the impending limit is shingled magnetic recording (...
A Universal SMR-aware Cache Framework with Deep Optimization for DM-SMR and HM-SMR Disks
To satisfy the enormous storage capacities required for big data, data centers have been adopting high-density shingled magnetic recording (SMR) disks. However, the weak fine-grained random write performance of SMR disks caused by their inherent write ...
RBC: A bandwidth controller to reduce write-stalls and tail latency
ICPP '23: Proceedings of the 52nd International Conference on Parallel ProcessingWe present RBC (Request Bandwidth Controller for LSM-tree), a controller that manages LSM-tree read and write operations and optimizes the requests for flash-based storage devices. RBC minimizes the number of write-stalls and reduces tail latency by ...
Comments