Data Center Management Research Articles

The Array Network Facility Seismic Bulletin: Products and an Unbiased View of United States Seismicity

Online Material: Movies of temporal deployment of USArray and seismicity: event processing flow chart; travel‐time and magnitude statistics plots; tables of earthquake catalogs, magnitude bias, and picked arrivals. The Array Network Facility (ANF), funded by the National Science Foundation Earthscope‐USArray project, is responsible for (1) collection of all transportable array (TA) station data and generation of station metadata; (2) their delivery to the Incorporated Research Institutions for Seismology Data Management Center (IRIS‐DMC). (3) Providing useful interfaces for field operations personnel at the Array Operations Facility (AOF) to obtain state‐of‐health information; (4) TA station command and control, such as seismometer mass centering and calibration; (5) collection of data from contributing regional network stations; and (6) data quality monitoring and control. Herein, we provide a brief overview of the evolution of USArray operations with emphasis on aspects relevant to the production of the ANF Seismic Bulletin. To begin operations at the ANF, real‐time acquisition of continuous three‐component broadband seismic data from 62 regional stations in California was initiated in April 2004. Seismic and state‐of‐health data from TA stations started flowing to the ANF shortly thereafter, with 73 stations transmitting data continuously by the end of 2004. A steady influx of data from newly deployed TA stations ensued with the first footprint of 400 USArray stations operating by mid‐September 2007 (Figs. 1a and 2a). USArray then began to migrate or roll eastward at an annual rate of about 200 stations, with the second and third distinct footprints being deployed by September 2009 and 2011, respectively (Fig. 1b and c). Deployment of the last distinct footprint was completed by October 2013. Figure 1d shows …

Benefits of a comprehensive quality program for cryopreserved PBMC covering 28 clinical trials sites utilizing an integrated, analytical web-based portal

The HIV Vaccine Trials Network (HVTN) is a global network of 28 clinical trial sites dedicated to identifying an effective HIV vaccine. Cryopreservation of high-quality peripheral blood mononuclear cells (PBMC) is critical for the assessment of vaccine-induced cellular immune functions. The HVTN PBMC Quality Management Program is designed to ensure that viable PBMC are processed, stored and shipped for clinical trial assays from all HVTN clinical trial sites. The program has evolved by developing and incorporating best practices for laboratory and specimen quality and implementing automated, web-based tools. These tools allow the site-affiliated processing laboratories and the central Laboratory Operations Unit to rapidly collect, analyze and report PBMC quality data. The HVTN PBMC Quality Management Program includes five key components: 1) Laboratory Assessment, 2) PBMC Training and Certification, 3) Internal Quality Control, 4) External Quality Control (EQC), and 5) Assay Specimen Quality Control. Fresh PBMC processing data is uploaded from each clinical site processing laboratory to a central HVTN Statistical and Data Management Center database for access and analysis on a web portal. Samples are thawed at a central laboratory for assay or specimen quality control and sample quality data is uploaded directly to the database by the central laboratory. Four year cumulative data covering 23,477 blood draws reveals an average fresh PBMC yield of 1.45×106±0.48cells per milliliter of useable whole blood. 95% of samples were within the acceptable range for fresh cell yield of 0.8–3.2×106cells/ml of usable blood. Prior to full implementation of the HVTN PBMC Quality Management Program, the 2007 EQC evaluations from 10 international sites showed a mean day 2 thawed viability of 83.1% and a recovery of 67.5%. Since then, four year cumulative data covering 3338 specimens used in immunologic assays shows that 99.88% had acceptable viabilities (>66%) for use in cellular assays (mean, 91.46% ±4.5%), and 96.2% had acceptable recoveries (50%–130%) with a mean of recovery of 85.8% ±19.12% of the originally cryopreserved cells. EQC testing revealed that since August 2009, failed recoveries dropped from 4.1% to 1.6% and failed viabilities dropped from 1.0% to 0.3%. The HVTN PBMC quality program provides for laboratory assessment, training and tools for identifying problems, implementing corrective action and monitoring for improvements. These data support the benefits of implementing a comprehensive, web-based PBMC quality program for large clinical trials networks.

Study of load balancing technology for EAST data management

With the continuous renewal and increasing number of diagnostics, the EAST tokamak routinely generates ∼3GB of raw data per pulse of the experiment, which is transferred to a centralized data management system. In order to strengthen international cooperation, all the acquired data has been converted and stored in the MDSplus servers. During the data system operation, there are some problems when a lot of client machines connect to a single MDSplus data server. Because the server process keeps the connection until the client closes its connection, a lot of server processes use a lot of network ports and consume a large amount of memory, so that the speed of access to data is very slow, but the CPU resource is not fully utilized. To improve data management system performance, many MDSplus servers will be installed on the blade server and form a server cluster to realize load balancing and high availability by using LVS and heartbeat technology. This paper will describe the details of the design and the test results.

Towards Scalable Traffic Management in Cloud Data Centers

Cloud Computing is becoming a mainstream paradigm, as organizations, large and small, begin to harness its benefits. This novel technology brings new challenges, mostly in the protocols that govern its underlying infrastructure. Traffic engineering in cloud data centers is one of these challenges that has attracted attention from the research community, particularly since the legacy protocols employed in data centers offer limited and unscalable traffic management. Many advocated for the use of VLANs as a way to provide scalable traffic management, however, finding the optimal traffic split between VLANs is the well known NP-Complete VLAN assignment problem. The size of the search space of the VLAN assignment problem is huge, even for small size networks. This paper introduce a novel decomposition approach to solve the VLAN mapping problem in cloud data centers through column generation. Column generation is an effective technique that is proven to reach optimality by exploring only a small subset of the search space. We introduce both an exact and a semi-heuristic decomposition with the objective to achieve load balancing by minimizing the maximum link load in the network. Our numerical results have shown that our approach explores less than 1% of the available search space, with an optimality gap of at most 4%. We have also compared and assessed the performance of our decomposition model and state of the art protocols in traffic engineering. This comparative analysis proves that our model attains encouraging gain over its peers.

Efficient and agile storage management in software defined environments

The IT industry is experiencing a disruptive trend for which the entire data center infrastructure is becoming software defined and programmable. IT resources are provisioned and optimized continuously according to a declarative and expressive specification of the workload requirements. The software defined environments facilitate agile IT deployment and responsive data center configurations that enable rapid creation and optimization of value-added services for clients. However, this fundamental shift introduces new challenges to existing data center management solutions. In this paper, we focus on the storage aspect of the IT infrastructure and investigate its unique challenges as well as opportunities in the emerging software defined environments. Current state-of-the-art software defined storage (SDS) solutions are discussed, followed by our novel framework to advance the existing SDS solutions. In addition, we study the interactions among SDS, software defined compute (SDC), and software defined networking (SDN) to demonstrate the necessity of a holistic orchestration and to show that joint optimization can significantly improve the effectiveness and efficiency of the overall software defined environments.

A Stochastic Model to Investigate Data Center Performance and QoS in IaaS Cloud Computing Systems

Cloud data center management is a key problem due to the numerous and heterogeneous strategies that can be applied, ranging from the VM placement to the federation with other clouds. Performance evaluation of cloud computing infrastructures is required to predict and quantify the cost-benefit of a strategy portfolio and the corresponding quality of service (QoS) experienced by users. Such analyses are not feasible by simulation or on-the-field experimentation, due to the great number of parameters that have to be investigated. In this paper, we present an analytical model, based on stochastic reward nets (SRNs), that is both scalable to model systems composed of thousands of resources and flexible to represent different policies and cloud-specific strategies. Several performance metrics are defined and evaluated to analyze the behavior of a cloud data center: utilization, availability, waiting time, and responsiveness. A resiliency analysis is also provided to take into account load bursts. Finally, a general approach is presented that, starting from the concept of system capacity, can help system managers to opportunely set the data center parameters under different working conditions.

Towards autonomic performance management of large scale data centers using interaction balance principle

In this paper, an autonomic performance management approach is introduced that can be applied to a general class of web services deployed in large scale distributed environment. The proposed approach utilizes traditional large scale control-based algorithms by using interaction balance approach in web service environment for managing the response time and the system level power consumption. This approach is developed in a generic fashion that makes it suitable for web service deployments, where web service performance can be adjusted by using a finite set of control inputs. This approach maintains the service level agreements, maximizes the revenue, and minimizes the infrastructure operating cost. Additionally, the proposed approach is fault-tolerant with respect to the failures of the computing nodes inside the distributed deployment. Moreover, the computational overhead of the proposed approach can also be managed by using appropriate value of configuration parameters during its deployment.

Optimal Power Allocation and Load Distribution for Multiple Heterogeneous Multicore Server Processors across Clouds and Data Centers

For multiple heterogeneous multicore server processors across clouds and data centers, the aggregated performance of the cloud of clouds can be optimized by load distribution and balancing. Energy efficiency is one of the most important issues for large-scale server systems in current and future data centers. The multicore processor technology provides new levels of performance and energy efficiency. The present paper aims to develop power and performance constrained load distribution methods for cloud computing in current and future large-scale data centers. In particular, we address the problem of optimal power allocation and load distribution for multiple heterogeneous multicore server processors across clouds and data centers. Our strategy is to formulate optimal power allocation and load distribution for multiple servers in a cloud of clouds as optimization problems, i.e., power constrained performance optimization and performance constrained power optimization. Our research problems in large-scale data centers are well-defined multivariable optimization problems, which explore the power-performance tradeoff by fixing one factor and minimizing the other, from the perspective of optimal load distribution. It is clear that such power and performance optimization is important for a cloud computing provider to efficiently utilize all the available resources. We model a multicore server processor as a queuing system with multiple servers. Our optimization problems are solved for two different models of core speed, where one model assumes that a core runs at zero speed when it is idle, and the other model assumes that a core runs at a constant speed. Our results in this paper provide new theoretical insights into power management and performance optimization in data centers.

GDCSim

Energy-efficient data center design and management has been a challenge of increasing importance in the past decade due to its potential to save billions of dollars in energy costs. However, the state of the art in design and evaluation of data centers require designers to be expertly familiar with a prohibitively large number of domain-specific design tools that necessitate user intervention in each step of the design process. This is due to the lack of a holistic data center design tool. To fill this gap, this article presents an iterative green data center design framework, the Green Data Center Simulator (GDCSim), for the design and development of energy-efficient data centers.

A Survey on Geographic Load Balancing Based Data Center Power Management in the Smart Grid Environment

Power management is becoming an increasingly important issue for Internet services supported by multiple geo-distributed data centers. These data center's energy consumptions and costs are becoming unacceptably high, and placing a heavy burden on both energy resources and the environment. Emerging smart grid provides a feasible way for dynamic and efficient power management of data centers. Various power management methodologies based on geographic load balancing (GLB) have recently been proposed to effectively utilize several features of smart grid. In this paper, we summarize the motivations, current state of the art, approaches and techniques proposed in the recent research works in this discipline. In all of these works, many perspectives of power management have been addressed using various computer science principles. We specifically elaborate on how researchers are exploiting mathematical tools to address these perspectives. Finally, we point out subject matters that need more attentions from the research community and provide our vision on possible future works along this direction.

PO-0933: A centralized data management system of heterogeneous data sources for radiation therapy

PO-0933: A centralized data management system of heterogeneous data sources for radiation therapy

Long-term transient thermal analysis using compact models for data center applications

Reduced-order thermal models are necessary to enable real-time assessment of the optimum operating and control conditions to improve data center energy efficiency. A 3D computational fluid dynamics (CFD) and heat transfer model of a basic raised floor, air cooled, hot aisle/cold aisle data center configuration was developed and simulation results were used to generate snapshots for initializing compact models based on the Proper Orthogonal Decomposition (POD) and the Nonlinear Principle Component Analysis (NLPCA) methods. The specific focus of the study was to numerically investigate how the thermal trends can be affected by the long-term transient flow patterns associated with leakage and how the compact models (e.g. POD and NLPCA) can be utilized for much faster implementations and characterizations of the transient flows. Using both the POD and the NLPCA method, good agreement was achieved between the full simulation results and the compact models for predicting the dynamically developed local flow structure over a range of transient cooling air supply operating conditions. In addition, the NLPCA method was implemented to better characterize the nonlinear aspects of the CFD results. The benefits of using both the POD and NLPCA methods are discussed in relation to constructing compact models as real-time predictive tools. A systematic use of the compact models has also been proposed, to enable more robust thermal management and control of data centers.

Adaptive Virtualized Resource Management for Application’s SLO Guarantees

在虚拟环境中实现应用服务级目标,是当前数据中心系统管理的关键问题之一.解决该问题有两个方面的要求:一方面,在虚拟化层次和范围内,能够动态和分布式地按需调整虚拟机资源分配;另一方面,在虚拟化范围之外,能够控制由于虚拟机对非虚拟化资源的竞争所导致的性能干扰,实现虚拟机性能隔离.然而,已有工作不适用于虚拟化数据中心场景.提出一种面向应用服务级目标的虚拟化资源管理方法.首先,该方法基于反馈控制理论,通过动态调整虚拟机资源分配来实现每个应用的服务器目标;同时,还设计了一个两层结构的自适应机制,使得应用模型能够动态地捕捉虚拟机资源分配与应用性能的时变非线性关系;最后,该方法通过仲裁不同应用的资源分配请求来控制虚拟机在非虚拟化资源上的竞争干扰.实验在基于Xen的机群环境中检验了该方法在RUBiS系统和TPC-W基准上的效果.实验结果显示,该方法的应用服务级目标实现率比两种对比方法平均高29.2%,而应用服务级目标平均偏离率比它们平均低50.1%.另一方面,当RUBiS系统和TPC-W基准竞争非虚拟化的磁盘I/O资源时,该方法通过抑制TPC-W基准28.7%的处理器资源需求来优先满足RUBiS系统的磁盘I/O需求.;Virtualized resources management for service level objectives (SLOs) of applications has been one of the key problems of system management in current data centers. To solve this problem one needs to: 1) dynamically and distributed allocating resources to virtual machines (VMs) of applications on demand; 2) efficiently control interference among VMs consolidated on a single physical server, due to their contention on non-virtualized resources. Many existing methods, however, are not suitable for this virtualized data center scenario. This paper presents a method for the virtualized resource management for SLOs of applications. First, based on the feedback control theory, this method can achieve SLOs of applications through dynamically resourced allocation. Second, a two-layer self-adaptive mechanism is devised and used to dynamically capture the non-linear relationship between the performance of applications and the resources allocation. Third, this method can control the performance interference among VMs on non-virtualized resources through virtualized resources allocation. The study has evaluated this method on the RUBiS system and TPC-W benchmark in a Xen-based virtualized cluster. The experimental results show that the average rate of SLOs achieved by this method is 29.2% higher than ones by two existing methods. Along with the average deviation from SLOs, this method is 50.1% lower than ones of the existing methods. Furthermore, when resource contention occurs on non-virtualized disk I/O between RUBiS and TPC-W, this method can almost entirely satisfy the disk I/O requirement of RUBiS of high priority through restraining TPC-W requests, e.g. 28.7%, on virtualized CPU.

Request–Response Distributed Power Management in Cloud Data Centers

AbstractPower provision is coming to be the most important constraint to data center development. The efficient management of power consumption according to the loads of the data center is urgent. As the load for every application hosted in every server node (SN) of the data center and corresponding Service Level Agreement (SLA) requirements can be quite different, it is hard to deploy a power strategy at application. The asynchronies and abruptness characteristics of workload fluctuation make power management policymaking using periodic resource scheduling method invalid. In this article, the design and implementation of the request–response distributed power management scheme is elaborated. Bound by linear time complexity, the method proposed integrates dynamic voltage/frequency scaling, power-on–power-off, and virtual machine migration mechanisms and dynamically optimizes the power consumption of a cloud data center. The significant advantage of the scheme is that it does not need synchronous scheduling between all SNs. Simulation results showed that the scheme could effectively decrease the power consumption of the data center, with a tiny reduction in performance as centralized control methods.

Service level agreement based energy-efficient resource management in cloud data centers

As cloud computing has become a popular computing paradigm, many companies have begun to build increasing numbers of energy hungry data centers for hosting cloud computing applications. Thus, energy consumption is increasingly becoming a critical issue in cloud data centers. In this paper, we propose a dynamic resource management scheme which takes advantage of both dynamic voltage/frequency scaling and server consolidation to achieve energy efficiency and desired service level agreements in cloud data centers. The novelty of the proposed scheme is to integrate timing analysis, queuing theory, integer programming, and control theory techniques. Our experimental results indicate that, compared to a statically provisioned data center that runs at the maximum processor speed without utilizing the sleep state, the proposed resource management scheme can achieve up to 50.3% energy savings while satisfying response-time-based service level agreements with rapidly changing dynamic workloads.

Recovering Host Overloading in Cloud using Virtual Machines

provides an efficient solution to the objectives of the cloud computing paradigm by facilitating creation of Virtual Machines (VMs) over the underlying physical servers, leading to improved resource utilization and intellection. Virtualization proposes to create a virtual version of a device or a resource likely to virtualize a server, a storage space, operating system or even network where the mechanism divides the resource into one or more execution environments. To analyze the behavior of a Cloud data center by means of various aspects like availability, utilization, responsiveness and waiting time. The factors that a cloud provider must take into account are elasticity, scalability, live migration of VMs and performance isolation. Live migration of VMs, the process of dynamically transferring a virtual machine across different servers, hasproved to represent a new opportunity to enable responsive and dynamic proved torepresent a new opportunity toenable responsive and dynamic resource management in modern data centers. Migrations of VMs is performed on the basis of overloading that occurs in physical servers in cloud data centers, whichleads to relatively less performance degradation. The SLA violation is also less compared to other techniques and this means the cloud provider will incur less cost from VM migrations. Further, MHOD algorithm performs host selection and VM reallocation quicker than the existing algorithms. Maintenance of physical servers can be efficiently achieved through our algorithm as it leads to efficient consolidation of VMs.

A model for collaborative clinical research in rare diseases: experience from the Rare Disease Clinical Research Network program

To overcome a lack of interest in research of rare diseases and an insufficient number of patients who can be recruited into clinical studies of rare diseases and potential diagnostics and treatments, the Rare Diseases Clinical Research Network (RDCRN) was established with the services of a centralized Data Management and Coordinating Center. The RDCRN has been very successful at recruiting patients, initiating clinical studies and trials, meeting patient recruitment milestones, training the next generation of research investigators and developing an active role for patient advocacy groups as research patnerships. The RDCRN has fostered international collaborations at multiple research sites that have led to regulatory approval of several drug products and expanded the knowledgebase for diseases included in the RDCRN. Rare disease research continues to evolve and considerable emphasis is placed on the translational research of discoveries from basic and clinical research activities, leading to the possibl...

The policy of free healthcare for children under the age of 6 years in Vietnam: assessment of the uptake for children hospitalised with acute diarrhoea in Ho Chi Minh City

ObjectiveTo assess the proportion of, and reasons for, households not utilising the policy of free healthcare for children under 6 years of age (FCCU6) for hospitalisation with diarrhoea, and assess the risk of catastrophic expenditure for households that forgo FCCU6 and pay out of pocket.MethodsInvoices detailing insurance information and charges incurred from 472 hospitalised diarrhoeal cases in one paediatric hospital in Ho Chi Minh City were retrieved. Hospital charges and the utilisation of elective services were analysed for patients utilising and not utilising FCCU6. Associations between socio-economic factors with non-utilisation of FCCU6 were evaluated.ResultsOverall, 29% of patients were FCCU6 non-users. The FCCU6 non-users paid a median hospital charge of $29.13 (interquartile range, IQR: $18.57–46.24), consuming no more than 1.4% of a medium-income household's annual income. Seventy per cent of low-income FCCU6 non-users utilised less-expensive elective services, whereas only 43% of medium income patients and 21% of high-income patients did (P = 0.036). Patients from larger households and those with a parent working in government were more likely to use FCCU6.ConclusionsThe rate of FCCU6 non-usage in this study population was 29%. A significant proportion of those that did not use FCCU6 was from lower income households and may perceive a justifiable cost–benefit ratio when forgoing FCCU6. Although a single diarrhoeal hospitalisation is unlikely to induce a catastrophic expenditure, FCCU6 non-usage may disproportionately increase the risk of catastrophic expenditure for lower income households over multiple illnesses.

Dashboard for Assessing Data Centers Green Compliance

Abstract: Nowadays, ICT sector activities and in particular Data Centers are known as an important environmental hazard. With the increasing popularity of the Internet and cloud computing, this threat seems to even get worse in the near future. Despite this increasing importance, there is still little have been done about data centers environmental affects and in particular measuring their green compliance level including all three Rs of waste management (Reuse, Reuse and Recycle). This paper tries to introduce a dashboard for evaluating data centers level of green compliance regardless of their tier. However, the dashboard is proposed based on Malaysias data centers condition, it still can be beneficial to data center managers in other parts of the world and researchers to open up new research possibilities.

A dynamic compact thermal model for data center analysis and control using the zonal method and artificial neural networks

Full-scale data center thermal modeling and optimization using computational fluid dynamics (CFD) is generally an extremely time-consuming process. This paper presents the development of a velocity propagation method (VPM) based dynamic compact zonal model to efficiently describe the airflow and temperature patterns in a data center with a contained cold aisle. Results from the zonal model are compared to those from full CFD simulations of the same configuration. A primary objective of developing the compact model is real-time predictive capability for control and optimization of operating conditions for energy utilization. A scheme is proposed that integrates zonal model results for temperature and air flow rates with a proportional–integral–derivative (PID) controller to predict and control rack inlet temperature more precisely. The approach also uses an Artificial Neural Network (ANN) in combination with a Genetic Algorithm (GA) optimization procedure. The results show that the combined approach, built on the VPM based zonal model, can yield an effective real-time design and control tool for energy efficient thermal management in data centers.