Data Center Efficiency Research Articles

Smart Monitoring Embedded Service for Energy-Efficient and Sustainable Management in Data Centers

Information technologies (IT) currently represent 2% of CO2 emissions. In recent years, a wide variety of IT solutions have been proposed, focused on increasing the energy efficiency of network data centers. Monitoring is one of the fundamental pillars of these systems, providing the information necessary for adequate decision making. However, today’s monitoring systems (MSs) are partial, specific and highly coupled solutions. This study proposes a model for monitoring data centers that serves as a basis for energy saving systems, offered as a value-added service embedded in a device with low cost and power consumption. The proposal is general in nature, comprehensive, scalable and focused on heterogeneous environments, and it allows quick adaptation to the needs of changing and dynamic environments. Further, a prototype of the system has been implemented in several devices, which has allowed validation of the proposal in addition to identification of the minimum hardware profile required to support the model.

Improvement of Energy Efficiency at Cloud Data Center Based on Fuzzy Markov Normal Algorithm VM Selection in Dynamic VM Consolidation

The problem of the energy consumption in cloud data center has become an important study in recent years. Dynamic VM consolidation has been known as a solution able to reduce energy consumption in cloud data center. However, there are many problems in Dynamic VM consolidation, in particular the VM selection. The objective of VM selection is to choose VM candidates suitable to move from overload host for avoiding oversubscribed host and to give an impact for the reduction of energy consumption. This research proposes a VM selection model in dynamic VM consolidation to improve the energy efficiency in cloud data center based on Fuzzy Markov Normal Algorithm. Fuzzy logic has been used for categorizing the attributes of VM candidates. After that, Markov Normal Algorithm has been used for deciding to which category VM should be migrated from overload host. The proposed VM selection model has been evaluated using various VM instance conditions (homogeneous or heterogeneous) with datasets from PlanetLabs in Cloudsim. Moreover, several parameters were used to measure the performance in this research such as Energy Consumption, SLA Violation, SLA Time per active host, and Performance Degradation Due Migration. The results experiments have shown the proposed VM selection model capable of improving energy efficiency in cloud data center up to 3.74%, 6.65%, 5.36%, and 5.11%, compared with the existing VM selections such as Constant First Selection, Minimum Migration Time, Random Choice, and Maximum Correlation.

Power management in virtualized datacenter – A survey

With growth in internet usage, increase in user demands, increase in applications or services due to technology advances (virtualization) and new set of business models (cloud computing), datacenters’ are growing in numbers and size. This datacenter growth, additionally contributes to increased energy consumption and thereby contributing to increased electricity cost and the concerns about the environmental impacts.Traditionally, computing systems were designed and developed to improve and meet performance demands of application or services. However, the increase in datacenter energy consumption and carbon footprint environmental impacts are real practical concerns. Therefore, the goal of the computing system design now is to optimally minimize power consumption and improve energy efficiency.This paper presents a survey of energy-efficient design of computing systems covering techniques used at hardware, virtualization, server heterogeneity and thermal levels. In this paper, we highlight the need to have a holistic view of power consumption of both IT and non-IT systems to achieve energy efficiency in datacenters.

Full-Stack Architecting to Achieve a Billion-Requests-Per-Second Throughput on a Single Key-Value Store Server Platform

Distributed in-memory key-value stores (KVSs), such as memcached, have become a critical data serving layer in modern Internet-oriented data center infrastructure. Their performance and efficiency directly affect the QoS of web services and the efficiency of data centers. Traditionally, these systems have had significant overheads from inefficient network processing, OS kernel involvement, and concurrency control. Two recent research thrusts have focused on improving key-value performance. Hardware-centric research has started to explore specialized platforms including FPGAs for KVSs; results demonstrated an order of magnitude increase in throughput and energy efficiency over stock memcached. Software-centric research revisited the KVS application to address fundamental software bottlenecks and to exploit the full potential of modern commodity hardware; these efforts also showed orders of magnitude improvement over stock memcached. We aim at architecting high-performance and efficient KVS platforms, and start with a rigorous architectural characterization across system stacks over a collection of representative KVS implementations. Our detailed full-system characterization not only identifies the critical hardware/software ingredients for high-performance KVS systems but also leads to guided optimizations atop a recent design to achieve a record-setting throughput of 120 million requests per second (MRPS) (167MRPS with client-side batching) on a single commodity server. Our system delivers the best performance and energy efficiency (RPS/watt) demonstrated to date over existing KVSs including the best-published FPGA-based and GPU-based claims. We craft a set of design principles for future platform architectures, and via detailed simulations demonstrate the capability of achieving a billion RPS with a single server constructed following our principles.

Guest Editors’ Introduction: Special Issue on Green and Energy-Efficient Cloud Computing: Part I

The papers in this special section focus on green and energy efficient cloud computing. Cloud computing has had a huge commercial impact and has attracted the interest of the research community. Public clouds allow their customers to outsource the management of physical resources, and rent a variable amount of resources in accordance to their specific needs. Private clouds allow companies to manage on-premises resources, exploiting the capabilities offered by the cloud technologies, such as using virtualization to improve resource utilization and cloud software for resource management automation. Hybrid clouds, where private infrastructures are integrated and complemented by external resources, are becoming a common scenario as well, for example to manage load peaks. Cloud applications are hosted by data centers whose size ranges from tens to tens of thousands of servers, which raises significant challenges related to energy and cost management. It has been estimated that the Information and Communication Technology (ICT) industry alone is responsible for 2-3 percent of the global greenhouse gas emissions. Therefore, we must find innovative methods and tools to manage the energy efficiency and carbon footprint of data centers, so that they can operate and scale in a cost-effective and environmentally sustainable manner. These methods and tools are often categorized as Data Center Infrastructure Management (DCIM) to monitor, control, and optimize data centers with extensive automation. DCIM must also effectively manage the quality of service provided by the data center, since cloud customers require high reliability, availability, usability, and low response times.

Goal-driven Optimization Strategy for Energy and Performance-Aware Data Centers for Cloud-Based Wind Farm CMS

A cloud computing system can be characterized by the provision of resources in the form of services to third parties on a leased, usage-based basis, as well as the private infrastructures maintained and utilized by individual organizations. To attain the desired reliability and energy efficiency in a cloud data center, trade-offs need to be carried out between system performance and power consumption. Resolving these conflicting goals is often the major challenge encountered in the design of optimization strategies for cloud data centers. The work presented in this paper is directed towards the development of an Energy-efficient and Performance-aware Cloud System equipped with strategies for dynamic switching of optimization approach. Moreover, a platform is also provided for the deployment of a Wind Farm CMS (Condition Monitoring System) which allows ubiquitous access. Due to the geographically-dispersed nature of wind farms, the CMS can take advantage of the cloud’s highly scalable architecture in order to keep a reliable and efficient operation capable of handling multiple simultaneous users and huge amount of monitoring data. Using the proposed cloud architecture, a Wind Farm CMS is deployed in a virtual platform to monitor and evaluate the aging conditions of the turbine’s major components in concurrent, yet isolated working environments.

Study of hot air recirculation and thermal management in data centers by using temperature rise distribution

One of the key challenges in reducing the energy consumption of data centers is to improve the efficiency of air conditioning units. The efficiency is increased when the temperature of the supply air rises; however, the hot air recirculation and hot spots may limit the rise of temperature of the supply air to ensure normal operation of servers. To solve this problem, the energy equation is modeled to separate the thermal influence of each server or rack, and the temperature rise distribution (TRD) is proposed to illustrate the recirculation efficiency in data centers. Based on the TRD, two recirculation ratios are suggested to evaluate the recirculation of each rack in the data center. An algorithm is suggested to optimize the power distribution among racks to minimize the maximum temperature inside the data center, reduce the number of hot spots, and provide more space for increasing the temperature of the supply air. The metrics and algorithm are tested and proved efficient by a computational fluid dynamics (CFD) case.

Providing a Cloud Broker-based Approach to Improve the Energy Consumption and Achieve a Green Cloud Computing

Today`s, cloud computing has been widely accepted in the industry. With the increasing popularity of cloud computing, many users and companies want to use and offer cloud computing services. The growth of cloud computing services may lead to consume a huge amount of energy and emit considerable amount of carbon dioxide. In recent years raising concerns about global warming and environment impact of greenhouse gases emission has led many researchers to engage in research in the field of green and energy aware computing. In this paper, “two phase carbon aware cloud broker” has been proposed that attempt to minimize energy and carbon by considering the energy and carbon efficiency of data centers. General Terms Cloud computing, carbon emissions, energy aware.

A Strategy to Improve the Efficiency of I/O Intensive Application in Cloud Computing Environment

I/O intensive application is one of the most popular cloud applications. However, few studies have focused on how to improve the processing efficiency of this specific type of application. Based on the previous studies on this kind of application, a package-based strategy is put forward. With this strategy, the files less than some threshold are packaged before the specific I/O operations. This can remarkably shorten the addressing and transmission time of these files and much improve the efficiency of I/O resources. To verify the strategy, we have done extensive experiments. The experimental results show the packaging strategy can efficiently improve the processing efficiency of I/O-intensive applications in cloud computing. This is beneficial for increasing the resource efficiency of cloud data centers.

A Survey of Virtual Machine Placement Techniques in a Cloud Data Center

Energy consumption of massive-scale cloud data centers is increasing unacceptably. There is a need to improve the energy efficiency of such data centers using Server Consolidation which aims at minimizing the number of Active Physical Machines (APMs) in a data center. Effective VM placement and migration techniques act as a key to optimum consolidation. Many of the recently proposed techniques realize dynamic consolidation while optimizing the VM placement. This paper presents a comprehensive study of the state-of-the-art VM placement and consolidation techniques used in green cloud which focus on improving the energy efficiency. A detailed comparison is presented, revealing pitfalls and suggesting improvisation methods along this direction.

A Big Data analyzer for large trace logs

Current generation of Internet-based services are typically hosted on large data centers that take the form of warehouse-size structures housing tens of thousands of servers. Continued availability of a modern data center is the result of a complex orchestration among many internal and external actors including computing hardware, multiple layers of intricate software, networking and storage devices, electrical power and cooling plants. During the course of their operation, many of these components produce large amounts of data in the form of event and error logs that are essential not only for identifying and resolving problems but also for improving data center efficiency and management. Most of these activities would benefit significantly from data analytics techniques to exploit hidden statistical patterns and correlations that may be present in the data. The sheer volume of data to be analyzed makes uncovering these correlations and patterns a challenging task. This paper presents Big Data analyzer (BiDAl), a prototype Java tool for log-data analysis that incorporates several Big Data technologies in order to simplify the task of extracting information from data traces produced by large clusters and server farms. BiDAl provides the user with several analysis languages (SQL, R and Hadoop MapReduce) and storage backends (HDFS and SQLite) that can be freely mixed and matched so that a custom tool for a specific task can be easily constructed. BiDAl has a modular architecture so that it can be extended with other backends and analysis languages in the future. In this paper we present the design of BiDAl and describe our experience using it to analyze publicly-available traces from Google data clusters, with the goal of building a realistic model of a complex data center.

A joint CPU-RAM energy efficient and SLA-compliant approach for cloud data centers

Cloud computing is a new paradigm that offers computing resources in a virtualized way with unprecedented levels of flexibility, reliability, and scalability. The benefits of cloud computing, however, come at a high cost in terms of energy consumption, mainly because of one of the cloud’s core enablers, the data center. There are a number of proposals that seek to enhance the energy efficiency of data centers. Still, most of them focus on the energy consumed by CPU and ignore other important hardware components, e.g., RAM. In this paper, we show the considerable impact that RAM can have on the total energy consumption, particularly in servers with large amounts of this memory. We then propose two new approaches for dynamic consolidation of virtual machines in cloud data centers that take into account both CPU and RAM usage. We have implemented and evaluated our proposals in the CloudSim simulator using real-world traces and compared the results with other state-of-the-art solutions. By adopting a wider view of the system, our proposals can reduce not only energy consumption but also service level agreement (SLA) violations, thus providing a better service at a lower cost.

Supporting Seamless Virtual Machine Migration via Named Data Networking in Cloud Data Center

Virtual machine migration has been touted as one of the crucial technologies in improving data center efficiency, such as reducing energy cost and maintaining load balance. However, traditional approaches could not avoid the service interruption completely. Moreover, they often result in longer delay and are prone to failures. In this paper, we leverage the emerging named data networking (NDN) to design an efficient and robust protocol to support seamless virtual machine migration in cloud data center. Specifically, virtual machines (VMs) are named with the services they provide. Request routing is based on service names instead of IP addresses that are normally bounded with physical machines. As such, services would not be interrupted when migrating supported VMs to different physical machines. We further analyze the performance of our proposed NDN-based VM migration protocol, and optimize its performance via a load balancing algorithm. Our extensive evaluations verify the effectiveness and the efficiency of our approach and demonstrate that it is interruption-free.

Investigation of Airflow Pattern of a Typical Data Center by CFD Simulation

Airflow pattern in a data center impacts the cooling effectiveness of the air-conditioning systems significantly, and hence impacts the energy efficiency of the whole data center. This paper presented investigation of the airflow pattern for a typical data center by CFD simulation. An under-floor air supplying system was used in the targeted data center. Cooling performance of the original and three more modified airflow pattern was simulated by CFD software. In optimizing the cooling efficiency, the measurements including blanking of the unused rack space, adding vertical partitions, and partly enclosing cold aisles were adopted to modify the original airflow pattern. The cooling performance was compared by indices like Supply Heat Index (SHI), Return Heat Index (RHI), Rack Cooling Index (RCI) and Return Temperature Index (RTI). The results indicated that partly enclosing cold aisles got most significant improvement in cooling performance. The supply air temperature could be increased by 3°C. While for the index itself, RTI was found not appropriate in evaluating of cold air by-pass or hot air re-circulation in some cases.

Power Efficient Hybrid VM Allocation Algorithm

Virtualization technology in Cloud computing has become important technology to reduce power consumption in data centers. Virtual Machine allocation to hosts is the main concept which carried out during Virtual Machine migrations in data centers. Virtual Machine allocation helps to utilize hardware resources of hosts and leads to power efficiency in Data centers. In the past few years, various mechanisms were proposed to apply algorithms to achieve power efficiency. In this paper, we have proposed a genetic algorithm to optimize various parameters i.e. power consumption, response time, SLA violation and VM migrations. Our proposed hybrid algorithm provisions various VMs to hosts in a way that to minimize power consumption, while delivering approved Quality of Service. Results demonstrate that proposed HVMA algorithm helps to minimize power consumption and to optimize various performance parameters during live migrations in various environment conditions.

Response Time Reduction and Performance Analysis of Load Balancing Algorithms at Peak Hours in Cloud Computing

are living in Digital Age which is undoubtedly the outcome of highly developed internet and its corresponding technologies. Cloud computing has become prominent as more people, organizations, entrepreneur, are moving towards it as it facilitates them to achieve their dreams in less investment. Now people from all over the globe are demanding for various services in rapid rate which has lead to bursty workloads on data centers thereby creating peak hour optimization problem to be handled. It has also lead to problem of load balancing which should be addressed to increase the efficiency of data centers. In the present work, peak hour optimization is being aimed and algorithms are being analyzed for large scale application to find efficient algorithm for real time scenario. Further, a novel strategy is being proposed to decrease the response time and DC processing time of algorithms without increasing the overall cost.

An adaptive fuzzy threshold-based approach for energy and performance efficient consolidation of virtual machines

Dynamic consolidation of virtual machines (VMs) is an effective technique, which can lead to improvement of energy efficiency and resource utilization in cloud data centers. However, due to varying...

Distributed consolidation of virtual machines for power efficiency in heterogeneous cloud data centers

Data centers use dynamic virtual machine consolidation to reduce power consumption. Existing consolidation mechanisms are not efficient in cloud data centers, which have heterogeneous hardware infrastructure, huge scale, and highly variable non-stationary workloads. We use game theory to develop a novel distributed mechanism for both heterogeneous and homogeneous data centers of cloud computing. Our mathematical analysis shows that our mechanism converges after a finite number of migrations. In addition, we show that our worst case power consumption is only 23% more than the theoretical minimum. In order to validate our claim, we preform simulation in CloudSim with real workload traces from Google data centers.

Time-Aware VMFlow Placement, Routing, and Migration for Power Efficiency in Data Centers

Increased power usage and network performance variation due to best-effort bandwidth sharing significantly affect tenancy cost, cloud adoption, and data center efficiencies. In this paper, we propose a novel time-aware request model which enables tenants to specify an estimated required time-duration, in addition to their required server resources for Virtual Machines (VMs) and network bandwidth for their communication. We investigate the VM-placement and routing problem, which allocates both server and network resources for the specified time-duration, to provide resource guarantees. Further, we exploit VM-migration while considering its power consumption overhead, to improve power saving and resource utilization. Using the multi-component utilization-based power model, we formulate the problem as an optimization problem that maximizes the acceptance rate while consuming as low power as possible. We develop fast online heuristics that allocate resources for requests, considering their duration and bandwidth demand. We also develop migration policies augmenting these heuristics. For migration heuristics, we propose server-migration and switch-migration approaches, which migrate the VMs between the powered-on servers only if their migrations result in turning-off at least one server and switch, respectively. We demonstrate the effectiveness of the proposed heuristics in terms of power saving, acceptance ratio, and migration overhead using comprehensive simulation results.

Architecting to achieve a billion requests per second throughput on a single key-value store server platform

Distributed in-memory key-value stores (KVSs), such as memcached, have become a critical data serving layer in modern Internet-oriented datacenter infrastructure. Their performance and efficiency directly affect the QoS of web services and the efficiency of datacenters. Traditionally, these systems have had significant overheads from inefficient network processing, OS kernel involvement, and concurrency control. Two recent research thrusts have focused upon improving key-value performance. Hardware-centric research has started to explore specialized platforms including FPGAs for KVSs; results demonstrated an order of magnitude increase in throughput and energy efficiency over stock memcached. Software-centric research revisited the KVS application to address fundamental software bottlenecks and to exploit the full potential of modern commodity hardware; these efforts too showed orders of magnitude improvement over stock memcached. We aim at architecting high performance and efficient KVS platforms, and start with a rigorous architectural characterization across system stacks over a collection of representative KVS implementations. Our detailed full-system characterization not only identifies the critical hardware/software ingredients for high-performance KVS systems, but also leads to guided optimizations atop a recent design to achieve a record-setting throughput of 120 million requests per second (MRPS) on a single commodity server. Our implementation delivers 9.2X the performance (RPS) and 2.8X the system energy efficiency (RPS/watt) of the best-published FPGA-based claims. We craft a set of design principles for future platform architectures, and via detailed simulations demonstrate the capability of achieving a billion RPS with a single server constructed following our principles.