{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T05:51:35Z","timestamp":1778133095648,"version":"3.51.4"},"reference-count":66,"publisher":"Association for Computing Machinery (ACM)","issue":"1","license":[{"start":{"date-parts":[[2022,1,29]],"date-time":"2022-01-29T00:00:00Z","timestamp":1643414400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"funder":[{"name":"National Key Research and Development Program of China","award":["2018YFA0701800"],"award-info":[{"award-number":["2018YFA0701800"]}]},{"DOI":"10.13039\/501100001809","name":"NSFC","doi-asserted-by":"crossref","award":["61821003 and 61872156"],"award-info":[{"award-number":["61821003 and 61872156"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100000001","name":"US NSF","doi-asserted-by":"crossref","award":["CCF-1704504 and CCF-1629625"],"award-info":[{"award-number":["CCF-1704504 and CCF-1629625"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Alibaba Group through Alibaba Innovative Research"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Storage"],"published-print":{"date-parts":[[2022,2,28]]},"abstract":"<jats:p>Hybrid storage servers combining solid-state drives (SSDs) and hard-drive disks (HDDs) provide cost-effectiveness and \u03bcs-level responsiveness for applications. However, observations from cloud storage system Pangu manifest that HDDs are often underutilized while SSDs are overused, especially under intensive writes. It leads to fast wear-out and high tail latency to SSDs. On the other hand, our experimental study reveals that a series of sequential and continuous writes to HDDs exhibit a periodic, staircase-shaped pattern of write latency, i.e., low (e.g., 35 \u03bcs), middle (e.g., 55 \u03bcs), and high latency (e.g., 12 ms), resulting from buffered writes within HDD\u2019s controller. It inspires us to explore and exploit the potential \u03bcs-level IO delay of HDDs to absorb excessive SSD writes without performance degradation.<\/jats:p>\n          <jats:p>\n            We first build an HDD writing model for describing the staircase behavior and design a profiling process to initialize and dynamically recalibrate the model parameters. Then, we propose a Buffer-Controlled Write approach (BCW) to proactively control buffered writes so that low- and mid-latency periods are scheduled with application data and high-latency periods are filled with padded data. Leveraging BCW, we design a mixed IO scheduler (MIOS) to adaptively steer incoming data to SSDs and HDDs. A multi-HDD scheduling is further designed to minimize HDD-write latency. We perform extensive evaluations under production workloads and benchmarks. The results show that MIOS removes up to 93% amount of data written to SSDs, reduces average and 99\n            <jats:sup>th<\/jats:sup>\n            -percentile latencies of the hybrid server by 65% and 85%, respectively.\n          <\/jats:p>","DOI":"10.1145\/3465410","type":"journal-article","created":{"date-parts":[[2022,1,29]],"date-time":"2022-01-29T19:07:01Z","timestamp":1643483221000},"page":"1-29","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":10,"title":["Exploration and Exploitation for Buffer-Controlled HDD-Writes for SSD-HDD Hybrid Storage Server"],"prefix":"10.1145","volume":"18","author":[{"given":"Shucheng","family":"Wang","sequence":"first","affiliation":[{"name":"Wuhan National Laboratory for Optoelectronics, HUST, Wuhan, Hubei, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ziyi","family":"Lu","sequence":"additional","affiliation":[{"name":"Wuhan National Laboratory for Optoelectronics, HUST, Wuhan, Hubei, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Qiang","family":"Cao","sequence":"additional","affiliation":[{"name":"Key Laboratory of Information Storage System of Ministry of Education, HUST, Wuhan, Hubei, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hong","family":"Jiang","sequence":"additional","affiliation":[{"name":"University of Texas at Arlington, Arlington, TX"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jie","family":"Yao","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, HUST, Wuhan, Hubei, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yuanyuan","family":"Dong","sequence":"additional","affiliation":[{"name":"Alibaba Group, Hangzhou, Zhejiang, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Puyuan","family":"Yang","sequence":"additional","affiliation":[{"name":"Alibaba Group, Hangzhou, Zhejiang, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Changsheng","family":"Xie","sequence":"additional","affiliation":[{"name":"Key Laboratory of Information Storage System of Ministry of Education, HUST, Wuhan, Hubei, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2022,1,29]]},"reference":[{"key":"e_1_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.23919\/DATE.2019.8714805"},{"key":"e_1_3_1_3_2","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2010.71"},{"key":"e_1_3_1_4_2","doi-asserted-by":"publisher","DOI":"10.1145\/1989323.1989414"},{"key":"e_1_3_1_5_2","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2018.00075"},{"key":"e_1_3_1_6_2","unstructured":"AXBOE. FIO: Flexible I\/O tester. Retrieved from https:\/\/github.com\/axboe\/fio."},{"key":"e_1_3_1_7_2","first-page":"753","volume-title":"Proceedings of the 2019 USENIX Annual Technical Conference (USENIX ATC\u201919)","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 (USENIX ATC\u201919). USENIX Association, Renton, WA, 753\u2013766. Retrieved fromhttps:\/\/www.usenix.org\/conference\/atc19\/presentation\/balmau."},{"key":"e_1_3_1_8_2","first-page":"115","volume-title":"Proceedings of the 8th USENIX Conference on File and Storage Technologies","author":"Boboila Simona","year":"2010","unstructured":"Simona Boboila and Peter Desnoyers. 2010. Write endurance in flash drives: Measurements and analysis. In Proceedings of the 8th USENIX Conference on File and Storage Technologies. 115\u2013128."},{"key":"e_1_3_1_9_2","volume-title":"Understanding the Linux Kernel: From I\/O Ports to Process Management","author":"Bovet Daniel P.","year":"2005","unstructured":"Daniel P. Bovet and Marco Cesati. 2005. Understanding the Linux Kernel: From I\/O Ports to Process Management. O\u2019Reilly Media, Inc."},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1145\/2043556.2043571"},{"key":"e_1_3_1_11_2","unstructured":"Alibaba Clouder. 2018. Pangu \u2013 the high performance distributed file system by Alibaba cloud. Retrieved from https:\/\/www.alibabacloud.com\/blog\/pangu_the_high_performance_distributed_file_system_by_alibaba_cloud_594059."},{"key":"e_1_3_1_12_2","unstructured":"Intel Corporation. 2016. Enterprise-class versus desktop-class hard drives. Retrieved from https:\/\/www.intel.com\/content\/dam\/support\/us\/en\/documents\/server-products\/Enterprise_vs_Desktop_HDDs_2.0.pdf."},{"key":"e_1_3_1_13_2","unstructured":"Intel Corporation. 2018. Product brief of intel optane solid state drive 905P. Retrieved from https:\/\/www.intel.com\/content\/dam\/www\/public\/us\/en\/documents\/product-briefs\/optane-ssd-905p-product-brief.pdf."},{"key":"e_1_3_1_14_2","unstructured":"Intel Corporation. 2019. Product brief of Intel 660p series. Retrieved from https:\/\/www.intel.com\/content\/dam\/www\/public\/us\/en\/documents\/product-briefs\/660p-series-brief.pdf."},{"key":"e_1_3_1_15_2","unstructured":"Western Digital Corporation. 2019. Product brief: WD gold enterprise class SATA HDD. Retrieved from https:\/\/documents.westerndigital.com\/content\/dam\/doc-library\/en_us\/assets\/public\/western-digital\/product\/internal-drives\/wd-gold\/product-brief-wd-gold-2579-810192.pdf."},{"key":"e_1_3_1_16_2","unstructured":"Western Digital Corporation. 2019. WD red NAS hard drives data sheet. Retrieved from http:\/\/products.wdc.com\/library\/SpecSheet\/ENG\/2879-800002.pdf."},{"key":"e_1_3_1_17_2","first-page":"205","volume-title":"ACM SIGOPS Operating Systems Review","author":"DeCandia Giuseppe","year":"2007","unstructured":"Giuseppe DeCandia, Deniz Hastorun, Madan Jampani, Gunavardhan Kakulapati, Avinash Lakshman, Alex Pilchin, Swaminathan Sivasubramanian, Peter Vosshall, and Werner Vogels. 2007. Dynamo: Amazon\u2019s highly available key-value store. ACM SIGOPS Operating Systems Review 41, 6 (2007), 205\u2013220."},{"key":"e_1_3_1_18_2","doi-asserted-by":"publisher","DOI":"10.1145\/3404397.3404437"},{"key":"e_1_3_1_19_2","doi-asserted-by":"publisher","DOI":"10.5555\/3323234.3323242"},{"key":"e_1_3_1_20_2","unstructured":"Samsung Electronics. 2017. Samsung SSD 960 EVO M.2 data sheet. Retrieved from https:\/\/www.intel.com\/content\/dam\/www\/public\/us\/en\/documents\/product-briefs\/660p-series-brief.pdf."},{"key":"e_1_3_1_21_2","doi-asserted-by":"publisher","DOI":"10.1145\/3093337.3037728"},{"key":"e_1_3_1_22_2","unstructured":"FUJITSU. 2007. MBC2073RC MBC2036RC hard disk drives product manual. 60\u201362. Retrieved from https:\/\/www.fujitsu.com\/downloads\/COMP\/fel\/support\/disk\/manuals\/c141-e266-01en.pdf."},{"key":"e_1_3_1_23_2","doi-asserted-by":"publisher","DOI":"10.1109\/MSST.2014.6855544"},{"key":"e_1_3_1_24_2","doi-asserted-by":"crossref","unstructured":"Sanjay Ghemawat Howard Gobioff and Shun-Tak Leung. 2003. The Google file system. ACM SIGOPS Operating Systems Review 37 5 (2003) 29\u201343.","DOI":"10.1145\/1165389.945450"},{"key":"e_1_3_1_25_2","doi-asserted-by":"crossref","unstructured":"Aayush Gupta Youngjae Kim and Bhuvan Urgaonkar. 2009. DFTL: A flash translation layer employing demand-based selective caching of page-level address mappings. ACM SIGPLAN Notices 44 3 (2009) 229\u2013240. ACM.","DOI":"10.1145\/1508284.1508271"},{"key":"e_1_3_1_26_2","doi-asserted-by":"publisher","DOI":"10.1145\/2517349.2522722"},{"key":"e_1_3_1_27_2","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2019.2897898"},{"key":"e_1_3_1_28_2","first-page":"289","volume-title":"ACM SIGARCH Computer Architecture News","author":"Jung Myoungsoo","year":"2014","unstructured":"Myoungsoo Jung, Wonil Choi, Shekhar Srikantaiah, Joonhyuk Yoo, and Mahmut T. Kandemir. 2014. HIOS: A host interface I\/O scheduler for solid state disks. ACM SIGARCH Computer Architecture News 42, 3 (2014), 289\u2013300."},{"key":"e_1_3_1_29_2","volume-title":"Proceedings of the 6th USENIX Workshop on Hot Topics in Storage and File Systems (HotStorage\u201914)","author":"Kang Jeong-Uk","year":"2014","unstructured":"Jeong-Uk Kang, Jeeseok Hyun, Hyunjoo Maeng, and Sangyeun Cho. 2014. The multi-streamed solid-state drive. In Proceedings of the 6th USENIX Workshop on Hot Topics in Storage and File Systems (HotStorage\u201914)."},{"key":"e_1_3_1_30_2","doi-asserted-by":"publisher","DOI":"10.1145\/1176760.1176774"},{"key":"e_1_3_1_31_2","doi-asserted-by":"publisher","DOI":"10.5555\/3358807.3358875"},{"key":"e_1_3_1_32_2","doi-asserted-by":"publisher","DOI":"10.5555\/3358807.3358875"},{"key":"e_1_3_1_33_2","doi-asserted-by":"publisher","DOI":"10.1145\/1629335.1629375"},{"key":"e_1_3_1_34_2","doi-asserted-by":"publisher","DOI":"10.5555\/3358807.3358876"},{"key":"e_1_3_1_35_2","doi-asserted-by":"publisher","DOI":"10.1145\/3208040.3208059"},{"key":"e_1_3_1_36_2","first-page":"501","volume-title":"Proceedings of the 2014 USENIX Annual Technical Conference","author":"Li Cheng","year":"2014","unstructured":"Cheng Li, Philip Shilane, Fred Douglis, Hyong Shim, Stephen Smaldone, and Grant Wallace. 2014. c: A capacity-optimized \\lbrace SSD \\rbrace cache for primary storage. In Proceedings of the 2014 USENIX Annual Technical Conference. 501\u2013512."},{"key":"e_1_3_1_37_2","doi-asserted-by":"publisher","DOI":"10.1145\/3302424.3303967"},{"key":"e_1_3_1_38_2","doi-asserted-by":"publisher","DOI":"10.1145\/3267809.3267830"},{"key":"e_1_3_1_39_2","doi-asserted-by":"publisher","DOI":"10.1145\/3357223.3362705"},{"key":"e_1_3_1_40_2","doi-asserted-by":"publisher","DOI":"10.1145\/2093139.2093143"},{"key":"e_1_3_1_41_2","first-page":"21","volume-title":"Proceedings of the Linux Symposium","volume":"2","author":"Mathur Avantika","year":"2007","unstructured":"Avantika Mathur, Mingming Cao, Suparna Bhattacharya, Andreas Dilger, Alex Tomas, and Laurent Vivier. 2007. The new ext4 filesystem: Current status and future plans. In Proceedings of the Linux Symposium. Vol. 2. 21\u201333."},{"key":"e_1_3_1_42_2","first-page":"1","volume-title":"Proceedings of the 10th USENIX conference on File and Storage Technologies","volume":"12","author":"Min Changwoo","year":"2012","unstructured":"Changwoo Min, Kangnyeon Kim, Hyunjin Cho, Sang-Won Lee, and Young Ik Eom. 2012. SFS: Random write considered harmful in solid state drives. In Proceedings of the 10th USENIX conference on File and Storage Technologies. Vol. 12. 1\u201316."},{"key":"e_1_3_1_43_2","first-page":"383","volume-title":"Proceedings of the 11th USENIX Symposium on Operating Systems Design and Implementation","author":"Muralidhar Subramanian","year":"2014","unstructured":"Subramanian Muralidhar, Wyatt Lloyd, Sabyasachi Roy, Cory Hill, Ernest Lin, Weiwen Liu, Satadru Pan, Shiva Shankar, Viswanath Sivakumar, Linpeng Tang, Sanjeev Kumar2014. f4: Facebook\u2019s warm BLOB storage system. In Proceedings of the 11th USENIX Symposium on Operating Systems Design and Implementation. 383\u2013398."},{"key":"e_1_3_1_44_2","doi-asserted-by":"publisher","DOI":"10.1109\/MSST.2011.5937225"},{"key":"e_1_3_1_45_2","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2014.86"},{"key":"e_1_3_1_46_2","doi-asserted-by":"publisher","DOI":"10.1145\/1383519.1383526"},{"key":"e_1_3_1_47_2","first-page":"13","volume-title":"Proceedings of the 10th USENIX conference on File and Storage Technologies","volume":"12","author":"Park Stan","year":"2012","unstructured":"Stan Park and Kai Shen. 2012. FIOS: A fair, efficient flash I\/O scheduler. In Proceedings of the 10th USENIX conference on File and Storage Technologies. Vol. 12. 13\u201313."},{"key":"e_1_3_1_48_2","doi-asserted-by":"publisher","DOI":"10.1145\/3035918.3056100"},{"key":"e_1_3_1_49_2","first-page":"101","volume-title":"Proceedings of the 8th USENIX Conference on File and Storage Technologies","volume":"10","author":"Soundararajan Gokul","year":"2010","unstructured":"Gokul Soundararajan, Vijayan Prabhakaran, Mahesh Balakrishnan, and Ted Wobber. 2010. Extending SSD lifetimes with disk-based write caches. In Proceedings of the 8th USENIX Conference on File and Storage Technologies. Vol. 10. 101\u2013114."},{"key":"e_1_3_1_50_2","article-title":"Linux block IO\u2013present and future","author":"Suse Jens","year":"2004","unstructured":"Jens Suse. 2004. Linux block IO\u2013present and future. In Proceedings of the Ottawa Linux Symposium.","journal-title":"Proceedings of the Ottawa Linux Symposium"},{"key":"e_1_3_1_51_2","first-page":"49","volume-title":"Proceedings of the 16th USENIX Conference on File and Storage Technologies","author":"Tavakkol Arash","year":"2018","unstructured":"Arash Tavakkol, Juan G\u00f3mez-Luna, Mohammad Sadrosadati, Saugata Ghose, and Onur Mutlu. 2018. MQSim: A framework for enabling realistic studies of modern multi-queue SSD devices. In Proceedings of the 16th USENIX Conference on File and Storage Technologies. Nitin Agrawal and Raju Rangaswami (Eds.). USENIX Association, 49\u201366."},{"key":"e_1_3_1_52_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2018.00041"},{"key":"e_1_3_1_53_2","unstructured":"Seagate Technology. 2016. Enhanced caching advantage\u2013TurboBoost and advanced write caching. Retrieved from https:\/\/www.seagate.com\/files\/www-content\/product-content\/enterprise-performance-savvio-fam\/enterprise-performance-15k-hdd\/_cross-product\/_shared\/doc\/enchanced-cache-advantage-tp691.1-1610us.pdf."},{"key":"e_1_3_1_54_2","unstructured":"Seagate Technology. 2018. Barracuda Pro Compute SATA HDD Data Sheet. Retrieved from https:\/\/www.seagate.com\/www-content\/datasheets\/pdfs\/barracuda-pro-14-tb-DS1901-9-1810US-en_US.pdf."},{"key":"e_1_3_1_55_2","unstructured":"Seagate Technology. 2019. Barracuda Compute SATA Product Manual. Retrieved from https:\/\/www.seagate.com\/www-content\/product-content\/desktop-hdd-fam\/en-us\/docs\/100799391e.pdf."},{"key":"e_1_3_1_56_2","doi-asserted-by":"publisher","DOI":"10.1109\/MSST.2013.6558426"},{"key":"e_1_3_1_57_2","doi-asserted-by":"publisher","DOI":"10.5555\/2591305.2591328"},{"key":"e_1_3_1_58_2","doi-asserted-by":"publisher","DOI":"10.1109\/EMSOFT.2013.6658584"},{"key":"e_1_3_1_59_2","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2018.00039"},{"key":"e_1_3_1_60_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N19-1242"},{"key":"e_1_3_1_61_2","doi-asserted-by":"publisher","DOI":"10.1145\/3121133"},{"key":"e_1_3_1_62_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-39527-7_5"},{"key":"e_1_3_1_63_2","doi-asserted-by":"publisher","DOI":"10.5555\/3357062.3357089"},{"key":"e_1_3_1_64_2","doi-asserted-by":"publisher","DOI":"10.1109\/MSST.2019.00-17"},{"key":"e_1_3_1_65_2","doi-asserted-by":"publisher","DOI":"10.1145\/1714454.1714456"},{"key":"e_1_3_1_66_2","first-page":"207","volume-title":"Proceedings of the 17th USENIX Conference on File and Storage Technologies","author":"Zheng Shengan","year":"2019","unstructured":"Shengan Zheng, Morteza Hoseinzadeh, and Steven Swanson. 2019. Ziggurat: A tiered file system for non-volatile main memories and disks. In Proceedings of the 17th USENIX Conference on File and Storage Technologies. 207\u2013219."},{"key":"e_1_3_1_67_2","doi-asserted-by":"crossref","unstructured":"Ke Zhou Si Sun Hua Wang Ping Huang Xubin He Rui Lan Wenyan Li Wenjie Liu and Tianming Yang. 2018. Demystifying cache policies for photo stores at scale: A Tencent case study. 284\u2013294. DOI:https:\/\/doi.org\/10.1145\/3205289.3205299","DOI":"10.1145\/3205289.3205299"}],"container-title":["ACM Transactions on Storage"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3465410","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3465410","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3465410","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,17]],"date-time":"2025-06-17T20:18:31Z","timestamp":1750191511000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3465410"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,1,29]]},"references-count":66,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2022,2,28]]}},"alternative-id":["10.1145\/3465410"],"URL":"https:\/\/doi.org\/10.1145\/3465410","relation":{},"ISSN":["1553-3077","1553-3093"],"issn-type":[{"value":"1553-3077","type":"print"},{"value":"1553-3093","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,1,29]]},"assertion":[{"value":"2020-10-01","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2021-05-01","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-01-29","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}