{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,18]],"date-time":"2026-01-18T00:31:31Z","timestamp":1768696291821,"version":"3.49.0"},"reference-count":65,"publisher":"Association for Computing Machinery (ACM)","issue":"4","content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Storage"],"published-print":{"date-parts":[[2025,11,30]]},"abstract":"<jats:p>\n                    Modern storage technologies aim at enhancing performance and lower costs. With advances in storage devices, numerous studies propose key-value store designs for heterogeneous storage systems. Many rely on the Log-Structured Merge-Tree (LSM), which optimizes write-heavy workloads and flash storage. However, LSM-tree-based key-value stores on heterogeneous storage systems suffer from severe performance degradation and a surge in query latency. A key issue is\n                    <jats:italic toggle=\"yes\">write stall constraints<\/jats:italic>\n                    , which arise from considering only single-tier storage, limiting key-value stores from leveraging multi-tier architectures. Another problem with key-value stores for heterogeneous storage systems is the\n                    <jats:italic toggle=\"yes\">inter-storage imbalance<\/jats:italic>\n                    . Performance discrepancies between storage tiers fluctuate due to each storage device\u2019s differing storage media technologies and garbage collection mechanisms. Consequently, existing\n                    <jats:italic toggle=\"yes\">hierarchical storage-based key-value stores<\/jats:italic>\n                    (HSKVS) do not fully utilize the resources of storage devices across tiers and do not account for the performance imbalance between storage tiers.\n                  <\/jats:p>\n                  <jats:p>\n                    This article presents an I\/O scheduler designed to maintain data balance across storage tiers of multi-tiered storage engines. Our approach leverages two key techniques:\n                    <jats:italic toggle=\"yes\">dynamic data layout<\/jats:italic>\n                    and\n                    <jats:italic toggle=\"yes\">flush I\/O throttling<\/jats:italic>\n                    .\n                    <jats:italic toggle=\"yes\">Dynamic data layout<\/jats:italic>\n                    optimizes data placement by adapting to performance metrics and workload demands. This approach ensures data is stored in the most appropriate tier, improving access times and reducing latency. Meanwhile,\n                    <jats:italic toggle=\"yes\">flush I\/O throttling<\/jats:italic>\n                    aims at efficiently managing the I\/O workload by controlling data movement between memory and storage. It balances data flow, preventing bottlenecks, especially during peak usage. To evaluate the effectiveness of our proposed I\/O scheduler, we integrated it into a conventional HSKVS system. This integration achieved up to a\n                    <jats:bold>14.7\u00d7<\/jats:bold>\n                    performance improvement on YCSB workload D compared to conventional HSKVS systems. Furthermore, our proposed scheduler reduced the tail latency for\n                    <jats:monospace>get<\/jats:monospace>\n                    queries by about\n                    <jats:bold>9.3\u00d7<\/jats:bold>\n                    , from 13 ms to 1.4 ms.\n                  <\/jats:p>","DOI":"10.1145\/3722224","type":"journal-article","created":{"date-parts":[[2025,3,8]],"date-time":"2025-03-08T06:42:46Z","timestamp":1741416166000},"page":"1-35","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Maintaining Inter-Layer Equilibrium in Hierarchical-Storage-based KV Store"],"prefix":"10.1145","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-1078-1772","authenticated-orcid":false,"given":"Kyoungho","family":"Koo","sequence":"first","affiliation":[{"name":"Electronic Engineering, Korea Advanced Institute of Science and Technology (KAIST)","place":["Daejeon, Korea (the Republic of)"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-5955-3555","authenticated-orcid":false,"given":"Junhan","family":"Lee","sequence":"additional","affiliation":[{"name":"Electrical Engineering, Korea Advanced Institute of Science and Technology","place":["Daejeon, Korea (the Republic of)"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2025,11,3]]},"reference":[{"key":"e_1_3_1_2_2","unstructured":"Facebook. 2021. RocksDB: Column Families. Retrieved from https:\/\/github.com\/facebook\/rocksdb\/wiki\/Column-Families"},{"key":"e_1_3_1_3_2","unstructured":"Google. 2024. LevelDB. Retrieved from https:\/\/github.com\/google\/leveldb. Accessed: 2025-03-17."},{"key":"e_1_3_1_4_2","unstructured":"Jens Axboe. 2024. FIO benchmark. Retrieved from https:\/\/fio.readthedocs.io\/en\/latest\/index.html. Accessed: 2025-03-17."},{"key":"e_1_3_1_5_2","unstructured":"Facebook. 2022. RocksDB: Rate Limiter. Retrieved from https:\/\/github.com\/facebook\/rocksdb\/wiki\/Rate-Limiter. Accessed: 2025-03-17."},{"key":"e_1_3_1_6_2","volume-title":"Proceedings of the 7th Symposium on Networked Systems Design and Implementation","author":"Anand Ashok","year":"2010","unstructured":"Ashok Anand, Chitra Muthukrishnan, Steven Kappes, Aditya Akella, and Suman Nath. 2010. Cheap and large CAMs for high performance data-intensive networked systems. In Proceedings of the 7th Symposium on Networked Systems Design and Implementation."},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1145\/2463676.2465296"},{"key":"e_1_3_1_8_2","unstructured":"Andrew Audibert. 2019. Scalable Metadata Service in Alluxio: Storing Billions of Files. Retrieved from https:\/\/www.alluxio.io\/blog\/scalable-metadata-service-in-alluxio-storing-billionsof-files\/. Accessed: 2025-03-17."},{"key":"e_1_3_1_9_2","volume-title":"Proceedings of the 2019 USENIX Annual Technical Conference","author":"Balmau Oana","year":"2019","unstructured":"Oana Balmau, Florin Dinu, Willy Zwaenepoel, Karan Gupta, Ravishankar Chandhiramoorthi, and Diego Didona. 2019. SILK: Preventing latency spikes in log-structured merge key-value stores. In Proceedings of the 2019 USENIX Annual Technical Conference."},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1145\/3380905"},{"key":"e_1_3_1_11_2","volume-title":"Proceedings of the 9th USENIX Symposium on Operating Systems Design and Implementation.","author":"Beaver Doug","year":"2010","unstructured":"Doug Beaver, Sanjeev Kumar, Harry C. Li, Jason Sobel, and Peter Vajgel. 2010. Finding a needle in haystack: Facebook\u2019s photo storage. In Proceedings of the 9th USENIX Symposium on Operating Systems Design and Implementation."},{"key":"e_1_3_1_12_2","doi-asserted-by":"publisher","DOI":"10.5555\/3386691.3386712"},{"key":"e_1_3_1_13_2","doi-asserted-by":"publisher","DOI":"10.5555\/3386691.3386712"},{"key":"e_1_3_1_14_2","first-page":"1007","volume-title":"Proceedings of the 2018 USENIX Annual Technical Conference","author":"Chan Helen H. W.","year":"2018","unstructured":"Helen H. W. Chan, Chieh-Jan Mike Liang, Yongkun Li, Wenjia He, Patrick P. C. Lee, Lianjie Zhu, Yaozu Dong, Yinlong Xu, Yu Xu, Jin Jiang, et\u00a0al. 2018. \\(\\lbrace\\) HashKV \\(\\rbrace\\) : Enabling efficient updates in \\(\\lbrace\\) KV \\(\\rbrace\\) storage via hashing. In Proceedings of the 2018 USENIX Annual Technical Conference. 1007\u20131019."},{"key":"e_1_3_1_15_2","doi-asserted-by":"publisher","DOI":"10.1145\/1365815.1365816"},{"key":"e_1_3_1_16_2","volume-title":"Proceedings of the 19th USENIX Conference on File and Storage Technologies.","author":"Chen Hao","year":"2021","unstructured":"Hao Chen, Chaoyi Ruan, Cheng Li, Xiaosong Ma, and Yinlong Xu. 2021. SpanDB: A fast, cost-effective LSM-tree based KV store on hybrid storage. In Proceedings of the 19th USENIX Conference on File and Storage Technologies."},{"key":"e_1_3_1_17_2","doi-asserted-by":"publisher","DOI":"10.1145\/2150976.2150998"},{"key":"e_1_3_1_18_2","doi-asserted-by":"publisher","DOI":"10.14778\/1454159.1454167"},{"key":"e_1_3_1_19_2","doi-asserted-by":"publisher","DOI":"10.1145\/1807128.1807152"},{"key":"e_1_3_1_20_2","doi-asserted-by":"publisher","DOI":"10.1145\/3183713.3196927"},{"key":"e_1_3_1_21_2","doi-asserted-by":"publisher","DOI":"10.1145\/2408776.2408794"},{"key":"e_1_3_1_22_2","doi-asserted-by":"publisher","DOI":"10.14778\/1920841.1921015"},{"key":"e_1_3_1_23_2","doi-asserted-by":"publisher","DOI":"10.1145\/1989323.1989327"},{"key":"e_1_3_1_24_2","doi-asserted-by":"publisher","DOI":"10.1145\/1323293.1294281"},{"key":"e_1_3_1_25_2","doi-asserted-by":"publisher","DOI":"10.1145\/3232559"},{"key":"e_1_3_1_26_2","doi-asserted-by":"publisher","DOI":"10.1145\/2367589.2367603"},{"key":"e_1_3_1_27_2","unstructured":"Krijn Doekemeijer and Animesh Trivedi. 2022. Key-value stores on flash storage devices: A survey. arXiv:2205.07975. Retrieved from https:\/\/arxiv.org\/abs\/2205.07975"},{"key":"e_1_3_1_28_2","doi-asserted-by":"publisher","DOI":"10.1145\/3483840"},{"key":"e_1_3_1_29_2","doi-asserted-by":"publisher","DOI":"10.1145\/3575693.3575715"},{"key":"e_1_3_1_30_2","doi-asserted-by":"publisher","DOI":"10.1145\/3190508.3190524"},{"key":"e_1_3_1_31_2","volume-title":"Proceedings of the Symposium on Mass Storage Systems and Technologies","author":"Factor Michael","year":"2005","unstructured":"Michael Factor, Kalman Meth, Dalit Naor, Ohad Rodeh, and Julian Satran. 2005. Object storage: The future building block for storage systems. In Proceedings of the Symposium on Mass Storage Systems and Technologies."},{"key":"e_1_3_1_32_2","volume-title":"Proceedings of the 12th USENIX Conference on File and Storage Technologies","author":"Harter Tyler","year":"2014","unstructured":"Tyler Harter, Dhruba Borthakur, Siying Dong, Amitanand Aiyer, Liyin Tang, Andrea C. Arpaci-Dusseau, and Remzi H. Arpaci-Dusseau. 2014. Analysis of HDFS under HBase: A Facebook messages case study. In Proceedings of the 12th USENIX Conference on File and Storage Technologies."},{"key":"e_1_3_1_33_2","doi-asserted-by":"publisher","DOI":"10.1145\/3299869.3314041"},{"key":"e_1_3_1_34_2","doi-asserted-by":"publisher","DOI":"10.1080\/00224065.1986.11979014"},{"key":"e_1_3_1_35_2","unstructured":"Junsu Im Jinwook Bae Chanwoo Chung Arvind and Sungjin Lee. 2020. PinK: High-speed in-storage key-value store with bounded tails. In Proceedings of the USENIX Annual Technical Conference (USENIX ATC\u201920). USENIX Association Berkeley CA 173\u2013187. https:\/\/www.usenix.org\/conference\/atc20\/presentation\/im"},{"key":"e_1_3_1_36_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.sysarc.2010.09.005"},{"key":"e_1_3_1_37_2","unstructured":"Intel. Optane Memory. Retrieved from https:\/\/www.intel.com\/content\/www\/us\/en\/products\/details\/memory-storage\/optane-memory.html. Accessed: 2025-03-17."},{"key":"e_1_3_1_38_2","unstructured":"Intel. 2018. Intel Optane SSD 905P. Retrieved from https:\/\/www.intel.co.kr\/content\/www\/kr\/ko\/products\/sku\/148607\/intel-optane-ssd-905p-series-380gb-m-2-110mm-pcie-x4-20nm-3d-xpoint\/specifications.html. Accessed: 2025-03-17."},{"key":"e_1_3_1_39_2","volume-title":"Proceedings of the USENIX Annual Technical Conference","author":"Kassa Hiwot Tadese","year":"2021","unstructured":"Hiwot Tadese Kassa, Jason Akers, Mrinmoy Ghosh, Zhichao Cao, Vaibhav Gogte, and Ronald Dreslinski. 2021. Improving performance of flash based key-value stores using storage class memory as a volatile memory extension. In Proceedings of the USENIX Annual Technical Conference."},{"key":"e_1_3_1_40_2","first-page":"183","volume-title":"Proceedings of the 13th USENIX Conference on File and Storage Technologies","author":"Kim Jaeho","year":"2015","unstructured":"Jaeho Kim, Donghee Lee, and Sam H. Noh. 2015. Towards \\(\\lbrace\\) SLO \\(\\rbrace\\) complying \\(\\lbrace\\) SSDs \\(\\rbrace\\) through \\(\\lbrace\\) OPS \\(\\rbrace\\) isolation. In Proceedings of the 13th USENIX Conference on File and Storage Technologies. 183\u2013189."},{"key":"e_1_3_1_41_2","doi-asserted-by":"publisher","DOI":"10.5626\/JOK.2022.49.7.561"},{"key":"e_1_3_1_42_2","doi-asserted-by":"publisher","DOI":"10.1109\/CCGrid.2014.51"},{"key":"e_1_3_1_43_2","doi-asserted-by":"publisher","DOI":"10.1109\/MSST.2015.7208288"},{"key":"e_1_3_1_44_2","doi-asserted-by":"publisher","DOI":"10.5626\/JCSE.2019.13.4.141"},{"key":"e_1_3_1_45_2","doi-asserted-by":"publisher","DOI":"10.1145\/3387902.3392621"},{"key":"e_1_3_1_46_2","first-page":"673","volume-title":"Proceedings of the 2021 USENIX Annual Technical Conference","author":"Li Yongkun","year":"2021","unstructured":"Yongkun Li, Zhen Liu, Patrick P. C. Lee, Jiayu Wu, Yinlong Xu, Yi Wu, Liu Tang, Qi Liu, and Qiu Cui. 2021. Differentiated \\(\\lbrace\\) Key-Value \\(\\rbrace\\) storage management for balanced \\(\\lbrace\\) I\/O \\(\\rbrace\\) performance. In Proceedings of the 2021 USENIX Annual Technical Conference. 673\u2013687."},{"key":"e_1_3_1_47_2","doi-asserted-by":"publisher","DOI":"10.1145\/3033273"},{"key":"e_1_3_1_48_2","unstructured":"C. Mellor. 2016. Toshiba Flashes 100TB QLC Flash Drive May go on Sale within Months. Really. Retrieved from https:\/\/www.theregister.com\/2016\/08\/10\/toshiba_100tb_qlc_ssd. Accessed: 2025-03-17."},{"key":"e_1_3_1_49_2","unstructured":"Micron. QLC NAND flash memory. Retrieved from https:\/\/www.micron.com\/products\/storage\/nand-flash\/qlc-nand. Accessed: 2025-03-17."},{"key":"e_1_3_1_50_2","unstructured":"Micron. 2020. What is a \u201cHeterogeneous Memory\u201d Storage Engine (HSE)? Retrieved from https:\/\/www.micron.com\/about\/blog\/company\/insights\/what-is-heterogeneous-memory-storage-engine-hse. Accessed: 2025-03-17."},{"key":"e_1_3_1_51_2","doi-asserted-by":"publisher","DOI":"10.1007\/s002360050048"},{"key":"e_1_3_1_52_2","doi-asserted-by":"publisher","DOI":"10.1145\/3582016.3582052"},{"key":"e_1_3_1_53_2","doi-asserted-by":"publisher","DOI":"10.1145\/3132747.3132765"},{"key":"e_1_3_1_54_2","unstructured":"Samsung. Samsung 870 QVO SSD. Retrieved from https:\/\/semiconductor.samsung.com\/consumer-storage\/internal-ssd\/870qvo\/. Accessed: 2025-03-17."},{"key":"e_1_3_1_55_2","unstructured":"Samsung. Samsung 970 EVO SSD. Retrieved from https:\/\/semiconductor.samsung.com\/consumer-storage\/internal-ssd\/970evo\/. Accessed: 2025-03-17."},{"key":"e_1_3_1_56_2","unstructured":"Samsung. 2018. Samsung Z-SSD SZ985. Retrieved from https:\/\/semiconductor.samsung.com\/news-events\/tech-blog\/samsung-z-ssd-sz985\/. Accessed: 2025-03-17."},{"key":"e_1_3_1_57_2","doi-asserted-by":"publisher","unstructured":"Subhadeep Sarkar Dimitris Staratzis Zichen Zhu and Manos Athanassoulis. 2021. Constructing and analyzing the LSM compaction design space. Proceedings of the VLDB Endowment 14 11 (2021) 2216\u20132229. 10.14778\/3476249.3476274","DOI":"10.14778\/3476249.3476274"},{"key":"e_1_3_1_58_2","doi-asserted-by":"publisher","DOI":"10.1145\/3575693.3575722"},{"key":"e_1_3_1_59_2","doi-asserted-by":"publisher","DOI":"10.5555\/2208461.2208479"},{"key":"e_1_3_1_60_2","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2020.3016680"},{"key":"e_1_3_1_61_2","volume-title":"Proceedings of the USENIX HotStorage","author":"Wu Kan","year":"2019","unstructured":"Kan Wu, Andrea Arpaci-Dusseau, and Remzi Arpaci-Dusseau. 2019. Towards an unwritten contract of intel optane SSD. In Proceedings of the USENIX HotStorage."},{"key":"e_1_3_1_62_2","doi-asserted-by":"publisher","DOI":"10.1145\/3472883.3487012"},{"key":"e_1_3_1_63_2","doi-asserted-by":"publisher","DOI":"10.1145\/3468521"},{"key":"e_1_3_1_64_2","volume-title":"Proceedings of the USENIX Annual Technical Conference","author":"Yao Ting","year":"2020","unstructured":"Ting Yao, Yiwen Zhang, Jiguang Wan, Qiu Cui, Liu Tang, Hong Jiang, Changsheng Xie, and Xubin He. 2020. MatrixKV: Reducing write stalls and write amplification in LSM-tree based KV stores with matrix container in NVM. In Proceedings of the USENIX Annual Technical Conference."},{"key":"e_1_3_1_65_2","volume-title":"Proceedings of the ACM Symposium on Cloud Computing","author":"Yoon Hobin","year":"2020","unstructured":"Hobin Yoon, Juncheng Yang, Sveinn Fannar Kristjansson, Steinn E. Sigurdarson, Ymir Vigfusson, and Ada Gavrilovska. 2020. Mutant: Balancing storage cost and latency in LSM-tree data stores. In Proceedings of the ACM Symposium on Cloud Computing."},{"key":"e_1_3_1_66_2","volume-title":"Proceedings of the 21st USENIX Conference on File and Storage Technologies","author":"Yu Jinghuan","year":"2023","unstructured":"Jinghuan Yu, Sam H. Noh, Young-ri Choi, and Chun Jason Xue. 2023. ADOC: Automatically harmonizing dataflow between components in log-structured key-value stores for improved performance. In Proceedings of the 21st USENIX Conference on File and Storage Technologies."}],"container-title":["ACM Transactions on Storage"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3722224","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,3]],"date-time":"2025-11-03T13:35:19Z","timestamp":1762176919000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3722224"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,3]]},"references-count":65,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2025,11,30]]}},"alternative-id":["10.1145\/3722224"],"URL":"https:\/\/doi.org\/10.1145\/3722224","relation":{},"ISSN":["1553-3077","1553-3093"],"issn-type":[{"value":"1553-3077","type":"print"},{"value":"1553-3093","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,3]]},"assertion":[{"value":"2023-12-04","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-02-14","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-11-03","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}