基于混合存储结构的分级协同节能方案

doi:10.3778/j.issn.1673-9418.2310065

摘要/Abstract

摘要： 新能源的应用已成为一种趋势，但具体到存储I/O领域，存储设备通常会根据存储负载的强度进行节点调度，而不规则波动的新能源和存储负载之间存在波峰波谷错位匹配的问题，增加了存储设备的调度难度。提出了一种针对NVM-SSD混合存储介质的分级协同节能方案MixSave，为每一层提供定制化节能方案并且协同调度，提升了新能源能效比。基于NVM和SSD存储介质的性能与能耗特点，MixSave采用NVM-SSD的高性价比分级存储架构，为用户提供高性能、大容量存储服务，并且通过副本保证数据可靠性。NVM作为缓存层采用负载驱动型方案，以保证性能为优先目标，即根据负载变化动态调整高功耗状态的NVM设备数量，满足负载需求；SSD作为数据层采用新能源驱动型节能方案，以提升新能源利用率为目标，根据新能源的变化动态调整SSD启用数量，进行缓存数据同步与预取。测试表明，与未采取节能措施的标准方案相比，MixSave性能下降不超过4%，在轻负载模式下，MixSave可节省73%~80%的传统能源，在重负载模式下，可节省55%~61%的传统能源。

关键词: 存储系统, 分级存储, 非易失性存储, 存储节能, 新能源

Abstract: Taking advantage of green energy is becoming increasingly popular. However, when it comes to the storage I/O, where storage devices usually schedule nodes based on the intensity of the storage workload, the irregular fluctuations of green energy lead to the mismatch of peaks and valleys between green energy and storage workloads, which increases the difficulty of scheduling storage resources. To address this issue, a tiering-cooperative energy-efficient scheme for hybrid NVM-SSD storage system, called MixSave, is proposed. MixSave customizes the energy-saving strategies for each layer and tunes them cooperatively between layers, reducing overall energy usage. It efficiently tiers NVM and SSD and replicates data among them based on their performance and energy characteristics, achieving both high performance and large capacity. The workload-driven strategy is used by the NVM layer, which is used for caching, to ensure the best performance, and the number of active NVM devices depends on the storage workload. Meanwhile, the SSD layer employs a green energy-oriented strategy to conserve traditional energy, and the number of active SSD devices will be proportionally to the amount of green energy supplied. Evaluation results show that, when compared with energy-save-unaware approaches, MixSave degrades performance by less than 4% and saves traditional energy by 73% to 80% and 55% to 61% under light and heavy workloads, respectively.

Key words: storage system, hierarchical storage, non-volatile memory, storage energy saving, green energy

李大平, 史庆宇, 唐忆滨, 胡哲琨, 高毅. 基于混合存储结构的分级协同节能方案[J]. 计算机科学与探索, 2024, 18(10): 2690-2703.

LI Daping, SHI Qingyu, TANG Yibin, HU Zhekun, GAO Yi. MixSave: Tiering-Cooperative Energy-Efficient Scheme for Hybrid Storage System[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(10): 2690-2703.

参考文献

[1] JONES N. How to stop data centres from gobbling up the world’s electricity[J]. Nature, 2018, 561(7722): 163-166.
[2] AUER H, CRESPO DEL GRANADO P, OEI P, et al. Erratum to: development and modelling of different decarbonization scenarios of the European energy system until 2050 as a contribution to achieving the ambitious 1.5℃ climate target-establishment of open source/data modelling in the European H2020 project openENTRANCE[J]. Elektrotechnik und Informationstechnik, 2021, 138(3): 256.
[3] MASANET E, SHEHABI A, LEI N, et al. Recalibrating global data center energy use estimates[J]. Science, 2020, 367(6481): 984-986.
[4] VERMA A, KOLLER R, USECHE L, et al. Srcmap: energy proportional storage using dynamic consolidation[C]//Proceedings of the 8th USENIX Conference on File and Storage Technologies, San Jose, Feb 23-26, 2010: 267-280.
[5] OH B, ABEYRATNE N, AHN J, et al. Enhancing DRAM self-refresh for idle power reduction[C]//Proceedings of the 2016 International Symposium on Low Power Electronics and Design, San Francisco, Aug 8-10, 2016. New York: ACM, 2016: 254-259.
[6] JIN T D, SHI T Q, PARK T. The quest for carbon-neutral industrial operations: renewable power purchase versus distributed generation[J]. International Journal of Production Research, 2018, 56(17): 5723-5735.
[7] 蔡浩然. 面向绿色数据中心的计算和能源高效协同调度方法研究[D]. 武汉: 华中科技大学, 2020.
CAI H R. Research on computing resources and power sources collaborative management approaches in green data centers[D]. Wuhan: Huazhong University of Science and Technology, 2020.
[8] KASSEM Y, CAMUR H, AATEG R A F. Exploring solar and wind energy as a power generation source for solving the electricity crisis in Libya[J]. Energies, 2020, 13(14): 3708-3719.
[9] LI C. Evaluation of the viability potential of four grid-connected solar photovoltaic power stations in Jiangsu province[J]. Clean Technologies and Environmental Policy, 2021, 23: 2117-2131.
[10] QU X Y, WAN J G, WANG J, et al. GreenMatch: renewable-aware workload scheduling for massive storage systems[C]//Proceedings of the 2016 IEEE International Parallel and Distributed Processing Symposium, Chicago, May 23-27, 2016. Washington: IEEE Computer Society, 2016: 403-412.
[11] ROSTIROLLA G, GRANGE L, MINH-THUYEN T, et al. A survey of challenges and solutions for the integration of renewable energy in datacenters[J]. Renewable and Sustainable Energy Reviews, 2022, 155: 111787.
[12] LIU Z H, CHEN Y, BASH C, et al. Renewable and cooling aware workload management for sustainable data centers[C]//Proceedings of the 2012 ACM SIGMETRICS/PERFORMANCE Joint International Conference on Measurement and Modeling of Computer Systems, London, Jun 11-15, 2012. New York: ACM, 2012: 175-186.
[13] BROWN M, RENAU J. ReRack: power simulation for data centers with renewable energy generation[J]. SIGMETRICS Performance Evaluation Review, 2011, 39(3): 77-81.
[14] GOIRI I, LE K, HAQUE M E, et al. GreenSlot: scheduling energy consumption in green datacenters[C]//Proceedings of the 2011 Conference on High Performance Computing Networking, Storage and Analysis, Seattle, Nov 12-18, 2011. New York: ACM, 2011.
[15] GOIRI I, LE K, NGUYEN T D, et al. GreenHadoop: leveraging green energy in data-processing frameworks[C]//Proceedings of the 7th EuroSys Conference 2012, Bern, Apr 10-13, 2012. New York: ACM, 2012: 57-70.
[16] GOIRI I, KATSAK W, LE K, et al. Parasol and GreenSwitch: managing datacenters powered by renewable energy[C]//Proceedings of the 2013 Architectural Support for Programming Languages and Operating Systems, Houston, Mar 16-20, 2013. New York: ACM, 2013: 51-64.
[17] KATSAK W, GOIRI I, BIANCHINI R, et al. GreenCassandra: using renewable energy in distributed structured storage systems[C]//Proceedings of the 6th International Green and Sustainable Computing Conference, Las Vegas, Dec 14-16, 2015. Washington: IEEE Computer Society, 2015: 1-8.
[18] WANG Q, CAI H R, CAO Q, et al. An energy-efficient power management for heterogeneous servers in data centers[J]. Computing, 2020, 102(7): 1717-1741.
[19] LI C, ZHANG W Y, CHO C B, et al. SolarCore: solar energy driven multi-core architecture power management[C]//Proceedings of the 17th International Conference on High-Performance Computer Architecture, San Antonio, Feb 12-16, 2011. Washington: IEEE Computer Society, 2011: 205-216.
[20] SHARMA N, BARKER S, IRWIN D, et al. Blink: managing server clusters on intermittent power[C]//Proceedings of the 16th International Conference on Architectural Support for Programming Languages and Operating Systems, Newport Beach, Mar 5-11, 2011. New York: ACM, 2011: 185-198.
[21] HAQUE M E, GOIRI I, BIANCHINI R, et al. GreenPar: scheduling parallel high performance applications in green datacenters[C]//Proceedings of the 2015 International Conference on Supercomputing, Newport Beach, Jun 8-11, 2015. New York: ACM, 2015: 217-227.
[22] LIU L J, LI C, SUN H B, et al. HEB: deploying and managing hybrid energy buffers for improving datacenter efficiency and economy[C]//Proceedings of the 42nd Annual International Symposium on Computer Architecture, Portland, Jun 13-17, 2015. New York: ACM, 2015: 463-475.
[23] LIU L J, SUN H B, LI C, et al. Exploring highly dependable and efficient datacenter power system using hybrid and hierarchical energy buffers[J]. IEEE Transactions on Sustainable Computing, 2021, 6(3): 412-426.
[24] SHEN L F, WANG F X, WANG F, et al. Backup battery allocation and workload migration against electrical load shedding at edge[J]. IEEE Internet of Things Journal, 2023, 10(19): 16804-16815.
[25] ZHOU X, CAI H R, CAO Q, et al. GreenGear: leveraging and managing server heterogeneity for improving energy efficiency in green data centers[C]//Proceedings of the 2016 International Conference on Supercomputing, Istanbul, Jun 1-3, 2016. New York: ACM, 2016.
[26] CAI H R, CAO Q, JIANG H, et al. GreenHetero: adaptive power allocation for heterogeneous green datacenters[C]//Proceedings of the 2021 IEEE 41st International Conference on Distributed Computing Systems, Washington, Jul 7-10, 2021. Piscataway: IEEE, 2021: 160-170.
[27] DENG N, STEWART C, LI J. Concentrating renewable energy in grid-tied datacenters[C]//Proceedings of the 2011 IEEE International Symposium on Sustainable Systems and Technology, Chicago, May 16-18, 2011. Piscataway: IEEE, 2011: 1-6.
[28] YAHOO! Inc. YAHOO! cloud serving benchmark in C++[EB/OL]. (2014-12-07) [2023-10-10]. https://github.com/basicthinker/YCSB-C.
[29] KHAN K N, HIRKI M, NIEMI T, et al. RAPL in action: experiences in using RAPL for power measurements[J]. ACM Transactions on Modeling and Performance Evaluation of Computing Systems, 2018, 3(2).
[30] National Renewable Energy Laboratory. Solar and wind datasets[EB/OL]. (2015-05-15) [2023-10-10]. https://www.nrel. gov/grid.