Energy Consumption Of Servers Research Articles

Sustainable Service Allocation Using a Metaheuristic Technique in a Fog Server for Industrial Applications

Reducing energy consumption in the fog computing environment is both a research and an operational challenge for the current research community and industry. There are several industries such as finance industry or healthcare industry that require a rich resource platform to process big data along with edge computing in fog architecture. As a result, sustainable computing in a fog server plays a key role in fog computing hierarchy. The energy consumption in fog servers depends on the allocation techniques of services (user requests) to a set of virtual machines (VMs). This service request allocation in a fog computing environment is a nondeterministic polynomial-time hard problem. In this paper, the scheduling of service requests to VMs is presented as a bi-objective minimization problem, where a tradeoff is maintained between the energy consumption and makespan. Specifically, this paper proposes a metaheuristic-based service allocation framework using three metaheuristic techniques, such as particle swarm optimization (PSO), binary PSO, and bat algorithm. These proposed techniques allow us to deal with the heterogeneity of resources in the fog computing environment. This paper has validated the performance of these metaheuristic-based service allocation algorithms by conducting a set of rigorous evaluations.

Simple estimation and energy‐aware migration models of virtual machines in a server cluster

SummaryIn order to realize green society, it is critical to reduce electric energy consumed by servers in clusters. In our previous studies, these types of algorithms are proposed to select an energy‐efficient server to perform an application process issued by a client. In this paper, we newly discuss a migration approach to reducing the electric energy consumption of servers where virtual machines with application processes migrate to more energy‐efficient servers. We propose a new algorithm called ISEAM2 to reduce the electric energy consumption of servers. Here, a pair of a host server and a virtual machine on the host server are first selected to perform a process issued by an application. In addition, a target virtual machine on a host server and a guest server to which the target virtual machine migrates are selected so that the electric energy consumption of the host and guest servers can be minimized. Thus, virtual machines with application processes migrate to more energy‐efficient servers. We also propose a simple way to estimate the termination time of every process on each server. In the evaluation, we show the total electric energy consumption and total active time of servers, and the average execution time of processes can be reduced in the ISEAM2 algorithm compared with other algorithms.

A cloud server energy consumption measurement system for heterogeneous cloud environments

With rapid development of cloud computing technologies and applications, the number and scale of cloud data centers have grown exponentially in recent years. One of the major problems with current cloud data centers is their huge energy consumption, which makes energy consumption management one of the hottest research topics in the field of cloud computing. This paper aims at implementing an effective Distributed Energy Meter (DEM) system for heterogeneous cloud environments based on a multi-component power consumption model for cloud servers. Specifically, we propose a modeling method for the energy consumption of key components (CPU, memory and disk) of computer servers and reveal the mathematical relationship between the resource usage of the key components and the system energy consumption. The proposed DEM system cannot only estimate the energy consumption of heterogeneous cluster environments (Linux and Windows NT), but also support various CPU power consumption models. In addition, a unique disk power consumption model that uses different thresholds to distinguish various disk I/O states (sequential/random, read/write) to achieve an accurate estimation of disk power consumption. Experimental studies conducted on a heterogeneous cluster with workloads generated by PCMark and Sysbench demonstrate that the proposed DEM system outperforms the state-of-art models in estimating the energy consumption of heterogeneous cloud environments.

Energy consumption laxity-based quorum selection for distributed object-based systems

In object based systems, an object is an unit of computation resource. Distributed applications are composed of multiple objects. Objects in an application are replicated to multiple servers in order to increase reliability, availability, and performance. On the other hand, the large amount of electric energy is consumed in a system compared with non-replication systems since multiple replicas of each object are manipulated on multiple servers. In this paper, the energy consumption laxity-based quorum selection (ECLBQS) algorithm is proposed to construct a quorum for each method issued by a transaction so that the total electric energy consumption of servers to perform methods can be reduced in the quorum based locking protocol. The total electric energy consumption of servers, the average execution time of each transaction, and the number of aborted transactions are shown to be more reduced in the ECLBQS algorithm than the random algorithm in evaluation.

Finding correlations of features affecting energy consumption and performance of web servers using the HADAS eco-assistant

The impact of energy consumption on the environment and the economy is raising awareness of “green” software engineering. HADAS is an eco-assistant that makes developers aware of the influence of their designs and implementations on the energy consumption and performance of the final product. In this paper, we extend HADAS to better support the requirements of users: researchers, automatically dumping the energy-consumption of different software solutions; and developers, who want to perform a sustainability analysis of different software solutions. This analysis has been extended by adding Pearson’s chi-squared differentials and Bootstrapping statistics, to automatically check the significance of correlations of the energy consumption, or the execution time, with any other variable (e.g., the number of users) that can influence the selection of a particular eco-efficient configuration. We have evaluated our approach by performing a sustainability analysis of the most common web servers (i.e. PHP servers) using the time and energy data measured with the Watts Up? Pro tool previously dumped in HADAS. We show how HADAS helps web server providers to make a trade-off between energy consumption and execution time, allowing them to sell different server configurations with different costs without modifying the hardware.

Dynamic Resource Caching in the IoT Application Layer for Smart Cities

We propose to apply constrained application protocol publish/subscribe to cache popular Internet of Things (IoT) resources in a broker to reduce the energy consumption of servers (which host these popular resources) and the average delay for delivering the IoT resources’ contents to the clients. We provide the smart parking application in smart cities as an example to demonstrate the benefit for conducting popular IoT resource caching. However, caching popular IoT resources in the broker may not always be the optimal choice, i.e., the broker may be congested by caching too many IoT resources, and so the average delay for enabling the broker to deliver contents of the IoT resources may be unbearable. Thus, we propose a novel energy aware and latency guaranteed dynamic resource caching (EASE) strategy to enable the broker to cache suitable popular resources such that the energy savings from the servers are maximized, while the average delay for publishing the contents of the resources to the corresponding clients is minimized. We demonstrate the performances of EASE via simulations as compared to other two baseline IoT resource caching strategies.

RAPL in Action

To improve energy efficiency and comply with the power budgets, it is important to be able to measure the power consumption of cloud computing servers. Intel’s Running Average Power Limit (RAPL) interface is a powerful tool for this purpose. RAPL provides power limiting features and accurate energy readings for CPUs and DRAM, which are easily accessible through different interfaces on large distributed computing systems. Since its introduction, RAPL has been used extensively in power measurement and modeling. However, the advantages and disadvantages of RAPL have not been well investigated yet. To fill this gap, we conduct a series of experiments to disclose the underlying strengths and weaknesses of the RAPL interface by using both customized microbenchmarks and three well-known application level benchmarks: Stream , Stress-ng , and ParFullCMS . Moreover, to make the analysis as realistic as possible, we leverage two production-level power measurement datasets from the Taito , a supercomputing cluster of the Finnish Center of Scientific Computing and also replicate our experiments on Amazon EC2. Our results illustrate different aspects of RAPL and document the findings through comprehensive analysis. Our observations reveal that RAPL readings are highly correlated with plug power, promisingly accurate enough, and have negligible performance overhead. Experimental results suggest RAPL can be a very useful tool to measure and monitor the energy consumption of servers without deploying any complex power meters. We also show that there are still some open issues, such as driver support, non-atomicity of register updates, and unpredictable timings that might weaken the usability of RAPL in certain scenarios. For such scenarios, we pinpoint solutions and workarounds.

Energy efficient job scheduling with workload prediction on cloud data center

Data centers are the backbone of cloud infrastructure platform to support large-scale data processing and storage. More and more business-to-consumer and enterprise applications are based on cloud data center. However, the amount of data center energy consumption is inevitably lead to high operation costs. The aim of this paper is to comprehensive reduce energy consumption of cloud data center servers, network, and cooling systems. We first build an energy efficient cloud data center system including its architecture, job and power consumption model. Then, we combine the linear regression and wavelet neural network techniques into a prediction method, which we call MLWNN, to forecast the cloud data center short-term workload. Third, we propose a heuristic energy efficient job scheduling with workload prediction solution, which is divided into resource management strategy and online energy efficient job scheduling algorithm. Our extensive simulation performance evaluation results clearly demonstrate that our proposed solution has good performance and is very suitable for low workload cloud data center.

Multiserver retrial queue with setup time and its application to data centers

This paper considers a multiserver retrial queue with setup time which is motivated from application in data centers with the ON-OFF policy, where an idle server is immediately turned off. The ON-OFF policy is designed to save energy consumption of idle servers because an idle server still consumes about 60% of its peak consumption processing jobs. Upon arrival, a job is allocated to one of available off-servers and that server is started up. Otherwise, if all the servers are not available upon arrival, the job is blocked and retries in a random time. A server needs some setup time during which the server cannot process a job but consumes energy. We formulate this model using a three-dimensional continuous-time Markov chain obtaining the stability condition via Foster-Lyapunov criteria. Interestingly, the stability condition is different from that of the corresponding non-retrial queue. Furthermore, exploiting the special structure of the Markov chain together with a heuristic technique, we develop an efficient algorithm for computing the stationary distribution. Numerical results reveal that under the ON-OFF policy, allowing retrials is more power-saving than buffering jobs. Furthermore, we obtain a new insight that if the setup time is relatively long, setting an appropriate retrial time could reduce both power consumption and the mean response time of jobs.

Dynamic migration of virtual machines to reduce energy consumption in a cluster

Virtual machines are widely used to support applications with virtual service in server clusters. Here, a virtual machine can migrate from a host server to a guest server. In this paper, we consider a cluster where virtual machines are dynamically created and dropped depending on the number of processes. We propose a dynamic virtual machine migration (DVMM) algorithm to reduce the total electric energy consumption of servers. If an application issues a process to a cluster, the most energy-efficient host server is first selected and then the process is performed on a virtual machine of the server. Then, a virtual machine migrates from a host server to a guest server so that total electric energy consumption of servers can be reduced. In the evaluation, we show the total electric energy consumption and active time of servers and the average execution time of processes can be reduced in the DVMM algorithm.

Simple algorithms for selecting an energy-efficient server in a cluster of servers

Electric energy consumed by computers, especially servers in scalable clusters has to be reduced to realise eco society. In this paper, we first propose a multi-level power consumption (MLPCM) model which gives how much electric power a server consumes to perform application processes. We also propose an multi-level computation (MLCM) model which shows the expected execution time of an application process performed on a server. By using the MLPCM and MLCM models, we propose a pair of simple locally energy-aware (SLEA) and simple globally energy-aware (SGEA) algorithms to select a host server to perform an application process issued by a client. In the SLEA algorithm, a host server is selected, which is expected to consume the minimum electric energy to perform the process. In the SGEA algorithm, a host server is selected, where the total expected electric energy consumption of not only the host server but also the other servers is minimum. In the evaluation, we show the total electric energy consumption and active time of servers and the average execution time of processes can be reduced in the SGEA algorithm compared with other algorithms.

A dynamic model for energy-efficiently migrating a group of virtual machines from a host server to a guest server

In this paper, we newly propose a dynamic migration of a group of υ virtual machines [DMG(υ)] algorithm to reduce the total electric energy consumption of servers in a cluster. Here, virtual machines are dynamically resumed and suspended so that the number of processes on each virtual machine can be kept fewer. In addition, a group of υ (υ ≥ 1) virtual machines migrate from a host server to a guest server so that the total electric energy consumed by the servers can be reduced. In our previous studies, time to migrate each virtual machine is assumed to be zero. In reality, there is migration time every process is suspended on a virtual machine migrating even in the live manner. Hence, the more often a virtual machine migrates, the longer time it takes to perform processes on the virtual machine. We propose a model to estimate electric energy to be consumed by servers through considering the migration time of each virtual machine. By using the estimation model, we can decide on which virtual machines on a host server to which guest server so that the total electric energy consumption of the host and guest servers can be reduced. In the evaluation, we show the total electric energy consumption of servers can be reduced by the DMG(υ) algorithm compared with other algorithms.

Measurement of Energy Performances for General-Structured Servers

Energy consumption of servers in data centers increases rapidly along with the wide application of Internet and connected devices. To improve the energy efficiency of servers, voluntary or mandatory energy efficiency programs for servers, including voluntary label program or mandatory energy performance standards have been adopted or being prepared in the US, EU and China. However, the energy performance of servers and testing methods of servers are not well defined. This paper presents matrices to measure the energy performances of general-structured servers. The impacts of various components of servers on their energy performances are also analyzed. Based on a set of normalized workload, the author proposes a standard method for testing energy efficiency of servers. Pilot tests are conducted to assess the energy performance testing methods of servers. The findings of the tests are discussed in the paper.

Desenvolvimento de solução ecoeficiente para data center

The survey evaluated the benefits, in terms of eco-efficiency, of IT virtualization from the main data center of the National Nuclear Energy Commission (CNEN). The eco-efficiency index chosen was the energy consumption of servers. The survey quantified CNEN's data center servers, and through exploration, identified the traditional servers, virtualization cluster servers, as well their respective virtual machines. The energy consumption of the servers was estimated using the tool Cost comparison toolkit: Calculating server power usage. The results show that all analyzed servers have low processing usage; the virtualization cluster of CNEN operates with excess resources on a scale that would allow the shutdown of two servers without compromising the operation of the cluster; a single cluster server is enough to house all non-virtualized servers in terms of processing and memory. The survey concludes that virtualization of traditional servers is the best way for CNEN to achieve a reduction in the energy consumption of its data center and that the virtualization cluster is wasting energy because it operates underutilized. This research contributes to a deeper look at virtualization as an alternative to data center management in public administration.

An energy-aware algorithm to migrate virtual machines in a server cluster

In order to realise eco-society, we have to reduce the electric energy consumed by servers in a cluster in addition to satisfying performance requirements. Virtual machines are now widely used to support applications with virtual computation service in clusters of servers. Here, a virtual machine on a host server can migrate to a guest server while processes are being performed on the virtual machine. In this paper, we propose an energy-aware migration of virtual machines (EAMV) algorithm where not only a virtual machine is selected to perform a process issued by a client but also a virtual machine migrates to a guest server which is expected to consume smaller electric energy to perform processes on the virtual machine. In the evaluation, we show the total electric energy consumption of servers and the average execution time of processes can be reduced in the EAMV algorithm compared with non-migration algorithms.

Multi-level power consumption model and energy-aware server selection algorithm

Electric power consumed by servers has to be reduced in a cluster in order to realise eco society. We take a macro-level approach to reducing the total electric energy consumption of servers to perform application processes. In this paper, we discuss a Multi-Level Power Consumption (MLPC) model of a server with a multi-thread CPU and a Multi-Level Computation (MLC) model. Based on the MLPC and the MLC models, we newly propose an energy-aware (EA) algorithm to select a server for each process requested by a client in a cluster to reduce the total electric energy consumption of the cluster while satisfying deadline requirements of the processes. We evaluate the EA algorithm and show that not only the total electric energy consumption of servers but also the average execution time of each process are reduced in the EA algorithm compared with the round-robin (RR) and random (RD) algorithms.

Optimizing the Energy Efficiency of the Modern Data Centers using Ant Colony Optimization

Objectives: To propose an energy efficient Swarm based Optimization method. Methods/Statistical Analysis: In this research, a set of tasks and servers are taken as Input. The data center server energy consumption is taken as the output of the algorithm. The processing time may vary according to the number of tasks given. Task allocation is done in such a way that most-efficient-server gets the tasks first. If, average job density is low, i.e. then it works on the same system and if the value is high then it moves to next one. The algorithm follows a scheduling and divides the jobs into servers and start execution of the job tasks. Findings: In this paper, an ant colony algorithm is proposed to efficiently allocate tasks to virtual machines, which allocates resources based on the available resources and the energy consumption of each virtual machine. This ACO algorithm is implemented, executed, and evaluated using the experiments in Cloud Sim. Reduced power consumption, with throughput has been obtained. Application/Improvements: Less SLA violation has been obtained with good response time.

Using Frequency Scaling on Virtualized Memory in Cloud Datacenters

As the increasing of IT-infrastructure in cloud platforms, rapidly growth of energy consumption becomes a critical problem in many cloud datacenters. Conventionally, most of studies on energy-efficiency optimization concentrate on CPU related energy costs instead of memory subsystem, since CPU often dominates the total energy consumption in modern servers. However, such a situation is gradually changing as more and more cloud datacenters are equipped with larger and larger memory systems for dealing with dataintensive applications. In this paper, we present a novel mechanism, namely frequency scaling on virtualized memory (FSVM), which applies DVFS technology on memory subsystem based on the characteristics of active VM instances. Comparing with previous studies, our approach provides a fine-grained memory energy consumption conservation mechanism for virtualized servers. Extensive experiments are conducted to investigate the effectiveness and performance of our FSVM, and the results indicate that it can significantly improve the energy-efficiency of memory subsystem in virtualized servers.

Optimal Web Page Download Scheduling Policies for Green Web Crawling

A web crawler is responsible for discovering and downloading new pages on the Web as well as refreshing previously downloaded pages. During these operations, the crawler issues a large number of HTTP requests to web servers. These requests increase the energy consumption and carbon footprint of the web servers since computational resources are used while serving the requests. In this work, we introduce the problem of green web crawling, where the objective is to devise a page refresh policy that minimizes the total staleness of pages in the repository of a web crawler, subject to a constraint on the amount of carbon emissions due to the processing on web servers. For the case of one web server and one crawling thread, the optimal policy turns out to be a greedy one. At each iteration, the page to be refreshed is selected based on a metric that considers the page's staleness, its size, and the greenness of the energy consumed at the web server premises. We then extend the optimal policy to the cases of 1) many servers; 2) multiple threads; and 3) pages with variable freshness requirements. We conduct simulations on a real data set that involves a large web server collection hosting around two billion pages. We present experimental results for the optimal page refresh policy as well as for various heuristics, in an effort to study the effect of different factors on performance.

New advances in large‐scale distributed simulation and real‐time applications

New advances in large‐scale distributed simulation and real‐time applications