The increasing integration of renewable energies into the electricity system and a fluctuating power demand cause high pressure on today's electricity grid. One contemporary approach to deal with this issue is the implementation of demand response mechanisms: In contrary to adjusting the power supply by increasing generation from brown energy sources, the demand side is rendered flexible. Data centers are highly suitable participants for this approach due to their extremely dynamic infrastructure and power consumption. Integration of data centers into demand response programs offers a huge potential and highly benefits energy providers, electricity market and grid. In this paper, we identify six basic theoretical compensation models from common demand response programs. Based on these, we provide an evaluation process, which is visualized as business process model with flowchart notation. This process supports the data center operator in evaluating and selecting a suitable demand response option.
{"title":"Evaluation process of demand response compensation models for data centers","authors":"Benedikt Kirpes, Sonja Klingert","doi":"10.1145/2940679.2940683","DOIUrl":"https://doi.org/10.1145/2940679.2940683","url":null,"abstract":"The increasing integration of renewable energies into the electricity system and a fluctuating power demand cause high pressure on today's electricity grid. One contemporary approach to deal with this issue is the implementation of demand response mechanisms: In contrary to adjusting the power supply by increasing generation from brown energy sources, the demand side is rendered flexible. Data centers are highly suitable participants for this approach due to their extremely dynamic infrastructure and power consumption. Integration of data centers into demand response programs offers a huge potential and highly benefits energy providers, electricity market and grid. In this paper, we identify six basic theoretical compensation models from common demand response programs. Based on these, we provide an evaluation process, which is visualized as business process model with flowchart notation. This process supports the data center operator in evaluating and selecting a suitable demand response option.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130697945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Increasing the use of renewable energy is considered a viable way of reducing carbon intensive power generation. However, a power grid running on high amounts of renewable energy has to deal with the limited controllability and higher volatility of power sources like wind or solar. In this work, we propose to use demand side management to deal with varying amounts of renewable power feed-in via the use of power plans, i.e. instructions passed to large energy consumers that specify how they should try to spread out their energy use over a day. We argue that a separation of power planning and implementation of technical measures to schedule loads to follow the plan would alleviate some of the problems faced by an integrated planning-scheduling approach, as these processes are governed by different entities who may be unwilling to disclose all required information to each other. As a proof-of-concept, we propose and analyze a quadratic programming approach to maximizing the fraction of renewable energy being used while not overburdening the consumer with a power plan that is difficult to follow.
{"title":"Energy supply aware power planning for flexible loads","authors":"F. Niedermeier, Fiodar Kazhamiaka, H. Meer","doi":"10.1145/2940679.2940681","DOIUrl":"https://doi.org/10.1145/2940679.2940681","url":null,"abstract":"Increasing the use of renewable energy is considered a viable way of reducing carbon intensive power generation. However, a power grid running on high amounts of renewable energy has to deal with the limited controllability and higher volatility of power sources like wind or solar. In this work, we propose to use demand side management to deal with varying amounts of renewable power feed-in via the use of power plans, i.e. instructions passed to large energy consumers that specify how they should try to spread out their energy use over a day. We argue that a separation of power planning and implementation of technical measures to schedule loads to follow the plan would alleviate some of the problems faced by an integrated planning-scheduling approach, as these processes are governed by different entities who may be unwilling to disclose all required information to each other. As a proof-of-concept, we propose and analyze a quadratic programming approach to maximizing the fraction of renewable energy being used while not overburdening the consumer with a power plan that is difficult to follow.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133613810","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Resource Virtualization and Load Balancing are main objectives to reduce the power consumption and to improve the performance of large data centers (DC). The management of Cloud Data Centers (CDC) requires an accurate planning and an efficient use of system resources in order to save energy consumption ("greening"), to provide Quality of Service (QoS), and to meet negotiated Service Level Agreements (SLA). This contribution addresses the question of modeling and the development of generic queuing models for energy-efficient use of resources for dynamic load balancing in virtualized CDCs. Performance models are developed for energy efficiency through automatic Server Consolidation, Dynamic Voltage and Frequency Scaling (DVFS) under Static Load Balancing; Dynamic Load Balancing can be achieved through Virtual Machine (VM) migrations. The analysis of such models provides quantitative performance figures upon which the system operation can be optimized with respect to guaranteed real-time performance and energy efficiency under prescribed SLAs.
{"title":"Energy effciency and performance of cloud data centers: which role can modeling play?","authors":"P. Kühn","doi":"10.1145/2940679.2940688","DOIUrl":"https://doi.org/10.1145/2940679.2940688","url":null,"abstract":"Resource Virtualization and Load Balancing are main objectives to reduce the power consumption and to improve the performance of large data centers (DC). The management of Cloud Data Centers (CDC) requires an accurate planning and an efficient use of system resources in order to save energy consumption (\"greening\"), to provide Quality of Service (QoS), and to meet negotiated Service Level Agreements (SLA). This contribution addresses the question of modeling and the development of generic queuing models for energy-efficient use of resources for dynamic load balancing in virtualized CDCs. Performance models are developed for energy efficiency through automatic Server Consolidation, Dynamic Voltage and Frequency Scaling (DVFS) under Static Load Balancing; Dynamic Load Balancing can be achieved through Virtual Machine (VM) migrations. The analysis of such models provides quantitative performance figures upon which the system operation can be optimized with respect to guaranteed real-time performance and energy efficiency under prescribed SLAs.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129781110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ariel Oleksiak, Wojciech Piątek, K. Kuczynski, Franciszek Sidorski
Data centers consume large amounts of energy. In parallel, power grid operators are struggling with reduction of peak energy demands. To cope with this problem automated demand response (ADR) techniques are developed. As data centers are precisely monitored and controlled they are well suited to the participation in ADR programmes. However, shapes of loads in data centres may be variable and not fitting the energy supply. We propose to apply ADR to data centers by adequate scheduling of workloads and the partial use of renewable energy sources (RES) especially during peak hours. As the energy produced by renewable sources may be very variable we investigate the use of energy storage. We present a model, heuristics that minimize overall energy cost and experimental results. We discuss profits for data centers that may come from participation in the demand response programme and the use of renewables energy sources.
{"title":"Reducing energy costs in data centres using renewable energy sources and energy storage","authors":"Ariel Oleksiak, Wojciech Piątek, K. Kuczynski, Franciszek Sidorski","doi":"10.1145/2940679.2940684","DOIUrl":"https://doi.org/10.1145/2940679.2940684","url":null,"abstract":"Data centers consume large amounts of energy. In parallel, power grid operators are struggling with reduction of peak energy demands. To cope with this problem automated demand response (ADR) techniques are developed. As data centers are precisely monitored and controlled they are well suited to the participation in ADR programmes. However, shapes of loads in data centres may be variable and not fitting the energy supply. We propose to apply ADR to data centers by adequate scheduling of workloads and the partial use of renewable energy sources (RES) especially during peak hours. As the energy produced by renewable sources may be very variable we investigate the use of energy storage. We present a model, heuristics that minimize overall energy cost and experimental results. We discuss profits for data centers that may come from participation in the demand response programme and the use of renewables energy sources.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131629284","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Geographical load balancing takes advantage of the regional differences in dynamic electricity rates by shifting computing tasks among geographically distributed data centers. Since energy storage is becoming an integral part of data centers, one can maximize the benefit of the temporal and spatial fluctuations of electricity rates by combining geographical load balancing and energy storage management. Previously, the problem of integrated geographical load balancing with energy storage has been studied based on Lyapunov stochastic optimization approach, which relies on asymptotic analysis by averaging over infinite time horizon and arbitrarily large energy storage. In this paper, we present a competitive online algorithmic approach, which can be applied to finite time horizon and small-to-medium energy storage with a worst-case guarantee from the offline optimal solutions. By simulations on real-world data, it is observed that our competitive online algorithms can significantly outperform Lyapunov optimization algorithm.
{"title":"Competitive online algorithms for geographical load balancing in data centers with energy storage","authors":"Chi-Kin Chau, L. Yang","doi":"10.1145/2940679.2940680","DOIUrl":"https://doi.org/10.1145/2940679.2940680","url":null,"abstract":"Geographical load balancing takes advantage of the regional differences in dynamic electricity rates by shifting computing tasks among geographically distributed data centers. Since energy storage is becoming an integral part of data centers, one can maximize the benefit of the temporal and spatial fluctuations of electricity rates by combining geographical load balancing and energy storage management. Previously, the problem of integrated geographical load balancing with energy storage has been studied based on Lyapunov stochastic optimization approach, which relies on asymptotic analysis by averaging over infinite time horizon and arbitrarily large energy storage. In this paper, we present a competitive online algorithmic approach, which can be applied to finite time horizon and small-to-medium energy storage with a worst-case guarantee from the offline optimal solutions. By simulations on real-world data, it is observed that our competitive online algorithms can significantly outperform Lyapunov optimization algorithm.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127124885","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
T. Cioara, I. Anghel, I. Salomie, Marcel Antal, M. Bertoncini, D. Arnone
The Data Centre (DC) services business is blooming increasing their energy demand and making them important players in the safe operation of the smart grid. The DC IT servers have a dynamic power factor varying from 0.98 for high density workloads and 0.8 for 20% server usage. A sub unitary power factor means that the voltage and current waveforms are not in phase making the electrical grid less efficient thus increasing the network loses. To make things even more challenging lately the grid operators have started to charge penalties if DC power factor is sub unitary. To improve their power factor with no extra hardware and no retrofitting, we propose an innovative method based on scheduling delay tolerant workload to achieve high servers' level usage thus increasing the leading factor and dynamically usage of nonelectrical cooling mechanisms such as the Thermal Energy Storage (TES) and this way increasing the lagging power factor. Simulation results are promising showing that a power factor with a value close to 1 can be achieved and maintained by proper planning the DC flexible power resources operation.
{"title":"Optimizing the power factor of data centers connected to the smart grid","authors":"T. Cioara, I. Anghel, I. Salomie, Marcel Antal, M. Bertoncini, D. Arnone","doi":"10.1145/2940679.2940682","DOIUrl":"https://doi.org/10.1145/2940679.2940682","url":null,"abstract":"The Data Centre (DC) services business is blooming increasing their energy demand and making them important players in the safe operation of the smart grid. The DC IT servers have a dynamic power factor varying from 0.98 for high density workloads and 0.8 for 20% server usage. A sub unitary power factor means that the voltage and current waveforms are not in phase making the electrical grid less efficient thus increasing the network loses. To make things even more challenging lately the grid operators have started to charge penalties if DC power factor is sub unitary. To improve their power factor with no extra hardware and no retrofitting, we propose an innovative method based on scheduling delay tolerant workload to achieve high servers' level usage thus increasing the leading factor and dynamically usage of nonelectrical cooling mechanisms such as the Thermal Energy Storage (TES) and this way increasing the lagging power factor. Simulation results are promising showing that a power factor with a value close to 1 can be achieved and maintained by proper planning the DC flexible power resources operation.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124242779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Data centers are among the largest and fastest growing consumers of electricity in the world. Furthermore, the rapid growth of digital content, big data, e-commerce, and internet traffic will create the need for an even higher number of DCs. On other side, in spite of the variability of renewable resources, due to characteristic weather fluctuations, the significant progress has been made in the renewable energy generation industry in terms of reducing installation cost and increasing integration into the power grid represents a good motive to tune data center software execution load in such a way that power consumption matches renewable energy availability (about 5%--30% of total DC load can be shifted [20]). This is especially viable in the context of smart cities, where the existence of a demand side management scheme can be assumed. In the context of the European project "DC4Cities", a similar scheme has been developed which consists of two phases. In the first phase, a concrete guidelines on power use for participating consumers to be followed is calculated. In the second phase, the control systems should find using this guidelines the best desired power values in terms of renewable percentage and SLAs. In this paper, an algorithm to calculate the aforementioned concrete guidelines by a component named "Max/Ideal Power Planner", based on smart city goals and renewable power availability forecasts, is proposed. In addition, the robustness of complete control system, particularly the Max/Ideal Power Planner, is estimated by evaluating the impact of renewable forecast accuracy on the scheduling of jobs in the data center via the proposed control system. Two types of errors in renewable forecasting are discussed: constant error and random error.
数据中心是世界上最大和增长最快的电力消费者之一。此外，数字内容、大数据、电子商务和互联网流量的快速增长将产生对更多数据中心的需求。另一方面，尽管可再生资源具有可变性，但由于典型的天气波动，可再生能源发电行业在降低安装成本和增加并入电网方面取得了重大进展，这是调整数据中心软件执行负载的良好动机，以使电力消耗与可再生能源可用性相匹配(约5%- 30%的总直流负载可以转移[20])。这在智慧城市的背景下尤其可行，因为可以假设存在需求侧管理方案。在欧洲项目“DC4Cities”的背景下，已经制定了一个类似的方案，该方案由两个阶段组成。在第一阶段，计算出参与消费者应遵守的具体用电准则。在第二阶段，控制系统应该根据可再生百分比和sla找到使用该指南所需的最佳功率值。本文提出了一种基于智慧城市目标和可再生能源可用性预测，通过“Max/Ideal Power Planner”组件计算上述具体指导方针的算法。此外，通过评估可再生预测精度对数据中心作业调度的影响，评估了整个控制系统，特别是Max/Ideal Power Planner的鲁棒性。讨论了可再生预测中的两种误差:恒定误差和随机误差。
{"title":"DC4Cities power planning: sensitivity to renewable energy forecasting errors","authors":"A. Alyousef, F. Niedermeier, H. Meer","doi":"10.1145/2940679.2940686","DOIUrl":"https://doi.org/10.1145/2940679.2940686","url":null,"abstract":"Data centers are among the largest and fastest growing consumers of electricity in the world. Furthermore, the rapid growth of digital content, big data, e-commerce, and internet traffic will create the need for an even higher number of DCs. On other side, in spite of the variability of renewable resources, due to characteristic weather fluctuations, the significant progress has been made in the renewable energy generation industry in terms of reducing installation cost and increasing integration into the power grid represents a good motive to tune data center software execution load in such a way that power consumption matches renewable energy availability (about 5%--30% of total DC load can be shifted [20]). This is especially viable in the context of smart cities, where the existence of a demand side management scheme can be assumed. In the context of the European project \"DC4Cities\", a similar scheme has been developed which consists of two phases. In the first phase, a concrete guidelines on power use for participating consumers to be followed is calculated. In the second phase, the control systems should find using this guidelines the best desired power values in terms of renewable percentage and SLAs. In this paper, an algorithm to calculate the aforementioned concrete guidelines by a component named \"Max/Ideal Power Planner\", based on smart city goals and renewable power availability forecasts, is proposed. In addition, the robustness of complete control system, particularly the Max/Ideal Power Planner, is estimated by evaluating the impact of renewable forecast accuracy on the scheduling of jobs in the data center via the proposed control system. Two types of errors in renewable forecasting are discussed: constant error and random error.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114727289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Modelling and analyzing power consumption for data centres can diagnose potential energy-hungry components and applications, and facilitate in-time control, benefiting the energy efficiency of data centers. However, solutions to this problem, including static power models and canonical prediction models, either aim to build a static relationship between power consumption and hardware/application configurations without considering the dynamic fluctuation of power; or simply treat it as time series, ignoring the inherit power data characteristics. To tackle these issues, in this paper, we present a systematic power prediction framework based on extensive power dynamic profiling and deep learning models. In particular, we first analyse different power series samples to illustrate their noise patterns; accordingly we propose a power data de-noising method, which lowers noise interference to the modelling. With the pretreated data, we propose two deep learning based prediction models, including a fine-grained model and a coarse-grained model, which are suitable for different time scales. In the fine-grained prediction model, a recursive autoencoder (AE) is employed for short-duration prediction; in the coarse-grained model, an AE is used to encode massive fine-grained historical data as a further data pretreatment for long-duration prediction. Experimental results show that our proposed models outperform canonical prediction methods with higher accuracy, up to 79% error reduction for certain cases.
{"title":"Learning-based power prediction for data centre operations via deep neural networks","authors":"Yuanlong Li, Han Hu, Yonggang Wen, Jun Zhang","doi":"10.1145/2940679.2940685","DOIUrl":"https://doi.org/10.1145/2940679.2940685","url":null,"abstract":"Modelling and analyzing power consumption for data centres can diagnose potential energy-hungry components and applications, and facilitate in-time control, benefiting the energy efficiency of data centers. However, solutions to this problem, including static power models and canonical prediction models, either aim to build a static relationship between power consumption and hardware/application configurations without considering the dynamic fluctuation of power; or simply treat it as time series, ignoring the inherit power data characteristics. To tackle these issues, in this paper, we present a systematic power prediction framework based on extensive power dynamic profiling and deep learning models. In particular, we first analyse different power series samples to illustrate their noise patterns; accordingly we propose a power data de-noising method, which lowers noise interference to the modelling. With the pretreated data, we propose two deep learning based prediction models, including a fine-grained model and a coarse-grained model, which are suitable for different time scales. In the fine-grained prediction model, a recursive autoencoder (AE) is employed for short-duration prediction; in the coarse-grained model, an AE is used to encode massive fine-grained historical data as a further data pretreatment for long-duration prediction. Experimental results show that our proposed models outperform canonical prediction methods with higher accuracy, up to 79% error reduction for certain cases.","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114777380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","authors":"","doi":"10.1145/2940679","DOIUrl":"https://doi.org/10.1145/2940679","url":null,"abstract":"","PeriodicalId":268208,"journal":{"name":"Proceedings of the 5th International Workshop on Energy Efficient Data Centres","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127327903","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：