Power optimization has become a key challenge in the design of large-scale enterprise data centers. Existing research efforts focus mainly on computer servers to lower their energy consumption, while only few studies have tried to address data center networks (DCNs), which can account for 10-20 percent of the total energy consumption of a data center. In this paper, we propose CARPO, a correlation-aware power optimization algorithm that dynamically consolidates traffic flows onto a small set of links and switches in a DCN and then shuts down unused network devices for energy savings. In sharp contrast to existing work, CARPO is designed based on a key observation from the analysis of real DCN traces that the bandwidth demands of different flows do not peak at exactly the same time. As a result, if the correlations among flows are considered in consolidation, more energy savings can be achieved. In addition, CARPO integrates traffic consolidation with link rate adaptation for maximized energy savings. We implement CARPO on a hardware testbed composed of 10 virtual switches configured with a production 48-port OpenFlow switch and 8 servers. Our empirical results with traces from Wikipedia and Yahoo! data centers demonstrate that CARPO can save up to 50 percent of network energy for a DCN, while having only negligible delay increases. CARPO also outperforms two state-of-the-art baselines by 19.6 and 95 percent on energy savings, respectively. Our simulation results with a large-scale DCN also show that CARPO can achieve more energy savings than the baselines for typical DCN topologies, such as fat tree and BCube.

Power optimization is a challenging task for a large scale data centre. Among all the major power consumption contributors of the entire data centre, data centre networks (DCNs) can account for up to 20% of the total power consumption. To save power for DCNs, several different approaches have been proposed. Unfortunately, none of these approaches addresses the performance impact on the flow completion times (FCTs) of delay-sensitive flows. In fact, FCTs of the delay-sensitive flows have been identified as one of the most important service-level agreement metrics of the data centre networking services. In this paper, we propose Power FCT, a multi-dimensional DCN power saving scheme that jointly conducts traffic consolidation and leverages the different power states of multiple components inside the DCN switches, including CPU, network processor, switch fabric, packet buffer and cooling fans, to maximize the power savings for DCNs. In sharp contrast to existing work that neglects the performance impacts, PowerFCT features a novel FCT model and explicitly enforces the desired FCT requirements in the process of power optimization. As a result, the desired FCTs for delay-sensitive flows can be effectively maintained. Both hardware experiments and simulation evaluations show that PowerFCT can achieve the power savings up to 62% for DCNs, which outperforms two state-of-the-art schemes by 21% and 14%, respectively. For large-scale data centers, this savings can lead to millions of dollars of savings in operating expenses. In addition, PowerFCT has the lowest FCT miss ratio among all the solutions.

Data centers have become a critical computing infrastructure in the era of cloud computing. Temperature monitoring and forecasting are essential for preventing server shutdowns because of overheating and improving a data center’s energy efficiency. This article presents a novel cyber-physical approach for temperature forecasting in data centers, one that integrates Computational Fluid Dynamics (CFD) modeling, in situ wireless sensing, and real-time data-driven prediction. To ensure forecasting fidelity, we leverage the realistic physical thermodynamic models of CFD to generate transient temperature distribution and calibrate it using sensor feedback. Both simulated temperature distribution and sensor measurements are then used to train a real-time prediction algorithm. As a result, our approach reduces not only the computational complexity of online temperature modeling and prediction, but also the number of deployed sensors, which enables a portable, noninvasive thermal monitoring solution that does not rely on the infrastructure of a monitored data center. We extensively evaluated the proposed system on a rack of 15 servers and a testbed of five racks and 229 servers in a small-scale production data center. Our results show that our system can predict the temperature evolution of servers with highly dynamic workloads at an average error of 0.52○C, within a duration up to 10 minutes. Moreover, our approach can reduce the required number of sensors by 67% while maintaining desirable prediction fidelity.

Data center power optimization has recently received a great deal of research attention. For example, server consolidation has been demonstrated as one of the most effective energy saving methodologies. Likewise, traffic consolidation has also been recently proposed to save energy for data center networks (DCNs). However, current research on data center power optimization focuses on servers and DCN separately. As a result, the optimization results are often inferior, because server consolidation without considering the DCN may cause traffic congestion and thus degraded network performance. On the other hand, server consolidation may change the DCN topology, allowing new opportunities for energy savings. In this paper, we propose PowerNetS, a power optimization strategy that leverages workload correlation analysis to jointly minimize the total power consumption of servers and the DCN. The design of PowerNetS is based on the key observations that the workloads of different servers and DCN traffic flows do not peak at exactly the same time. Thus, more energy savings can be achieved if the workload correlations are considered in server and traffic consolidations. In addition, PowerNetS considers the DCN topology during server consolidation, which leads to less inter-server traffic and thus more energy savings and shorter network delays. We implement PowerNetS on a hardware testbed composed of 10 virtual switches configured with a production 48-port OpenFlow switch and 6 servers. Our empirical results with Wikipedia, Yahoo!, and IBM traces demonstrate that PowerNetS can save up to 51.6% of energy for a data center. PowerNetS also outperforms two state-of-the-art baselines by 44.3% and 15.8% on energy savings, respectively. Our simulation results with 72 switches and 122 servers also show the superior energy efficiency of PowerNetS over the baselines.

It is well known that periodically putting nodes into sleep can effectively save energy in wireless sensor networks at the cost of increased communication delays. However, most existing work mainly focuses on the static sleep scheduling, which cannot guarantee the desired delay when the network conditions change dynamically. In many applications with user-specified end-to-end delay requirements, the duty cycle of every node should be tuned individually at runtime based on the network conditions to achieve the desired end-to-end delay guarantees and energy efficiency. In this article, we propose DutyCon, a control theory-based dynamic duty-cycle control approach. DutyCon decomposes the end-to-end delay guarantee problem into a set of single-hop delay guarantee problems along each data flow in the network. We then formulate the single-hop delay guarantee problem as a dynamic feedback control problem and design the controller rigorously, based on feedback control theory, for analytic assurance of control accuracy and system stability. DutyCon also features a queuing delay adaptation scheme that adapts the duty cycle of each node to unpredictable incoming packet rates, as well as a novel energy-balancing approach that extends the network lifetime by dynamically adjusting the delay requirement allocated to each hop. Our empirical results on a hardware testbed demonstrate that DutyCon can effectively achieve the desired trade-off between end-to-end delay and energy conservation. Extensive simulation results also show that DutyCon outperforms two baseline sleep scheduling protocols by having more energy savings while meeting the end-to-end delay requirements.

Multi-channel communications can effectively reduce channel competition and interference in a wireless sensor network and thus achieve increased throughput and improved end-to-end delay guarantees with reduced power consumption. However, existing work relies only on a small number of orthogonal channels, resulting in degraded performance when a large number of data flows need to be transmitted on different channels. In this article, we conduct empirical studies to investigate interferences among overlapping channels. Our results show that overlapping channels can also be utilized for improved real-time performance if the node transmission power is carefully configured. In order to minimize the overall transmission power consumption of a network with multiple data flows under end-to-end delay constraints, we formulate a constrained optimization problem to configure the transmission power level of every node and assign overlapping channels to different data flows. Since the optimization problem has an exponential computational complexity, we then present a heuristic algorithm designed based on simulated annealing to find a suboptimal solution. Our extensive empirical results on a 33-mote testbed demonstrate that our algorithm achieves better real-time performance and less power consumption than two baselines, including a scheme using only orthogonal channels.

Recent studies have shown that a significant portion of the total energy consumption of many data centers is caused by the inefficient operation of their cooling systems. Without effective thermal monitoring with accurate location information, the cooling systems often use unnecessarily low temperature set points to overcool the entire room, resulting in excessive energy consumption. Sensor network technology has recently been adopted for data-center thermal monitoring because of its nonintrusive nature for the already complex data center facilities and robustness to instantaneous CPU or disk activities. However, existing solutions place sensors in a simplistic way without considering the thermal dynamics in data centers, resulting in unnecessarily degraded hot server detection probability. In this paper, we first formulate the problems of sensor placement for hot server detection in a data center as constrained optimization problems in two different scenarios. We then propose a novel placement scheme based on computational fluid dynamics (CFD) to take various factors, such as cooling systems and server layout, as inputs to analyze the thermal conditions of the data center. Based on the CFD analysis in various server overheating scenarios, we apply data fusion and advanced optimization techniques to find a near-optimal sensor placement solution, such that the probability of detecting hot servers is significantly improved. Our empirical results in a real server room demonstrate the detection performance of our placement solution. Extensive simulation results in a large-scale data center with 32 racks also show that the proposed solution outperforms several commonly used placement solutions in terms of detection probability.

Power optimization has become a key challenge in the design of large-scale enterprise data centers. Existing research efforts focus mainly on computer servers to lower their energy consumption, while only few studies have tried to address the energy consumption of data center networks (DCNs), which can account for 20% of the total energy consumption of a data center. In this paper, we propose CARPO, a correlation-aware power optimization algorithm that dynamically consolidates traffic flows onto a small set of links and switches in a DCN and then shuts down unused network devices for energy savings. In sharp contrast to existing work, CARPO is designed based on a key observation from the analysis of real DCN traces that the bandwidth demands of different flows do not peak at exactly the same time. As a result, if the correlations among flows are considered in consolidation, more energy savings can be achieved. In addition, CARPO integrates traffic consolidation with link rate adaptation for maximized energy savings. We implement CARPO on a hardware testbed composed of 10 virtual switches configured with a production 48-port OpenFlow switch and 8 servers. Our empirical results with Wikipedia traces demonstrate that CARPO can save up to 46% of network energy for a DCN, while having only negligible delay increases. CARPO also outperforms two state-of-the-art baselines by 19.6% and 95% on energy savings, respectively. Our simulation results with 61 flows also show the superior energy efficiency of CARPO over the baselines.

Data centers have become a critical computing infrastructure in the era of cloud computing. Temperature monitoring and forecasting are essential for preventing overheating-induced server shutdowns and improving a data center's energy efficiency. This paper presents a novel cyber-physical approach for temperature forecasting in data centers, which integrates Computational Fluid Dynamics (CFD) modeling, in situ wireless sensing, and real-time data-driven prediction. To ensure the forecasting fidelity, we leverage the realistic physical thermodynamic models of CFD to generate transient temperature distribution and calibrate it using sensor feedback. Both simulated temperature distribution and sensor measurements are then used to train a real-time prediction algorithm. As a result, our approach significantly reduces the computational complexity of online temperature modeling and prediction, which enables a portable, noninvasive thermal monitoring solution that does not rely on the infrastructure of monitored data center. We extensively evaluated our system on a rack of 15 servers and a test bed of five racks and 229 servers in a production data center. Our results show that our system can predict the temperature evolution of servers with highly dynamic workloads at an average error of 0.52C, within a duration up to 10 minutes.

Recent studies have shown that a significant portion of the total energy consumption of many data centers is caused by the inefficient operation of their cooling systems. Without effective thermal monitoring with accurate location information, the cooling systems often use unnecessarily low temperature set points to over cool the entire room, resulting in excessive energy consumption. Sensor network technology has recently been adopted for data-center thermal monitoring because of its non-intrusive nature for the already complex data center facilities and robustness to instantaneous CPU or disk activities. However, existing solutions place sensors in a simplistic way without considering the thermal dynamics in data centers, resulting in unnecessarily degraded hot server detection probability. In this paper, we first formulate the problem of sensor placement for hot server detection in a data center as a constrained optimization problem. We then propose a novel placement scheme based on Computational Fluid Dynamics (CFD) to take various factors, such as cooling systems and server layout, as inputs to analyze the thermal conditions of the data center. Based on the CFD analysis in various server overheating scenarios, we apply data fusion and advanced optimization techniques to find a near-optimal sensor placement solution, such that the probability of detecting hot servers is significantly improved. Our empirical results in a real server room demonstrate the detection performance of our placement solution. Extensive simulation results also show that the proposed solution outperforms a commonly used placement solution in terms of detection probability.

As many radio chips used in today’s sensor mote hardware can work at different frequencies, several multichannel communication protocols have recently been proposed to improve network throughput and reduce packet loss for wireless sensor networks. However, existing work cannot utilize multiple channels to provide explicit guarantees for application-specified end-to-end communication delays, which are critical to many real-time applications such as surveillance and disaster response. In this article, we propose MCRT, a multichannel real-time communication protocol that features a flow-based channel allocation strategy. Because of the small number of orthogonal channels available in current mote hardware, MCRT allocates channels to network partitions formed based on many-to-one data flows. To achieve bounded end-to-end communication delay for every data flow, the channel allocation problem has been formulated as a constrained optimization problem and proven to be NP-complete. We then present the design of MCRT, which includes a channel allocation algorithm and a real-time packet forwarding strategy. Extensive simulation results based on a realistic radio model and empirical results on a real hardware testbed of Tmote nodes both demonstrate that MCRT can effectively utilize multiple channels to reduce the number of deadlines missed in end-to-end communications. Our results also show that MCRT outperforms a state-of-the-art real-time protocol and two baseline multichannel communication schemes.

It is well known that periodically putting nodes into sleep can effectively save energy in wireless sensor networks, at the cost of increased communication delays. However, most existing work mainly focuses on static sleep scheduling, which cannot guarantee the desired delay when the network conditions change dynamically. In many applications with user-specified end-to-end delay requirements, the duty cycle of every node should be tuned individually at runtime based on the network conditions to achieve the desired end-to-end delay guarantees and energy efficiency. In this paper, we propose DutyCon, a control theory-based dynamic duty cycle control approach. DutyCon decomposes the end-to-end delay guarantee problem into a set of single-hop delay guarantee problems along each data flow in the network. We then formulate the single-hop delay guarantee problem as a dynamic feedback control problem and design the controller rigorously, based on feedback control theory, for analytic assurance of control accuracy and system stability. DutyCon also features a queuing delay adaptation scheme that adapts the duty cycle of each node to unpredictable packet rates, as well as a novel energy balancing approach that extends the network lifetime by dynamically adjusting the delay requirement allocated to each hop. Our empirical results on a hardware testbed demonstrate that DutyCon can effectively achieve the desired tradeoff between end-to-end delay and energy conservation. Extensive simulation results also show that DutyCon outperforms two baseline sleep scheduling protocols by having more energy savings while meeting the end-to-end delay requirements.