scispace - formally typeset
Search or ask a question

Showing papers by "Violeta Holmes published in 2015"


01 Jan 2015
TL;DR: In this paper, the authors present the results of their work in deploying Docker containers in the cluster environment and an evaluation of its suitability as a runtime for high performance parallel execution.
Abstract: Linux container technology has more than proved itself useful in cloud computing as a lightweight alternative to virtualisation, whilst still offering good enough resource isolation. Docker is emerging as a popular runtime for managing Linux containers, providing both management tools and a simple file format. Research into the performance of containers compared to traditional Virtual Machines and bare metal shows that containers can achieve near native speeds in processing, memory and network throughput. A technology born in the cloud, it is making inroads into scientific computing both as a format for sharing experimental applications and as a paradigm for cloud based execution. However, it has unexplored uses in traditional cluster and grid computing. It provides a run time environment in which there is an opportunity for typical cluster and parallel applications to execute at native speeds, whilst being bundled with their own specific (or legacy) library versions and support software. This offers a solution to the Achilles heel of cluster and grid computing that requires the user to hold intimate knowledge of the local software infrastructure. Using Docker brings us a step closer to more effective job and resource management within the cluster by providing both a common definition format and a repeatable execution environment. In this paper we present the results of our work in deploying Docker containers in the cluster environment and an evaluation of its suitability as a runtime for high performance parallel execution. Our findings suggest that containers can be used to tailor the run time environment for an MPI application without compromising performance, and would provide better Quality of Service for users of scientific computing.

37 citations


Book ChapterDOI
12 Jul 2015
TL;DR: The findings suggest that containers can be used to tailor the run time environment for an MPI application without compromising performance, and would provide better Quality of Service for users of scientific computing.
Abstract: Linux container technology has more than proved itself useful in cloud computing as a lightweight alternative to virtualisation, whilst still offering good enough resource isolation. Docker is emerging as a popular runtime for managing Linux containers, providing both management tools and a simple file format. Research into the performance of containers compared to traditional Virtual Machines and bare metal shows that containers can achieve near native speeds in processing, memory and network throughput. A technology born in the cloud, it is making inroads into scientific computing both as a format for sharing experimental applications and as a paradigm for cloud based execution. However, it has unexplored uses in traditional cluster and grid computing. It provides a run time environment in which there is an opportunity for typical cluster and parallel applications to execute at native speeds, whilst being bundled with their own specific (or legacy) library versions and support software. This offers a solution to the Achilles heel of cluster and grid computing that requires the user to hold intimate knowledge of the local software infrastructure. Using Docker brings us a step closer to more effective job and resource management within the cluster by providing both a common definition format and a repeatable execution environment. In this paper we present the results of our work in deploying Docker containers in the cluster environment and an evaluation of its suitability as a runtime for high performance parallel execution. Our findings suggest that containers can be used to tailor the run time environment for an MPI application without compromising performance, and would provide better Quality of Service for users of scientific computing.

35 citations


Journal ArticleDOI
01 Jan 2015
TL;DR: The experience in developing the HPC, Cluster and Grid modules is presented including a review of existing HPC courses offered at the UK universities and suggestions for future work are made.
Abstract: High-Performance Computing (HPC) and the ability to process large amounts of data are of paramount importance for UK business and economy as outlined by Rt Hon David Willetts MP at the HPC and Big Data conference in February 2014. However there is a shortage of skills and available training in HPC to prepare and expand the workforce for the HPC and Big Data research and development. Currently, HPC skills are acquired mainly by students and staff taking part in HPC-related research projects, MSc courses, and at the dedicated training centres such as Edinburgh Universitys EPCC. There are few UK universities teaching the HPC, Clusters and Grid Computing courses at the undergraduate level. To address the issue of skills shortages in the HPC it is essential to provide teaching and training as part of both postgraduate and undergraduate courses. The design and development of such courses is challenging since the technologies and software in the fields of large scale distributed systems such as Cluster, Cloud and Grid computing are undergoing continuous change. The students completing the HPC courses should be proficient in these evolving technologies and equipped with practical and theoretical skills for future jobs in this fast developing area. In this paper we present our experience in developing the HPC, Cluster and Grid modules including a review of existing HPC courses offered at the UK universities. The topics covered in the modules are described, as well as the coursework project based on practical laboratory work. We conclude with an evaluation based on our experience over the last ten years in developing and delivering the HPC modules on the undergraduate courses, with suggestions for future work.

17 citations


Journal ArticleDOI
16 May 2015
TL;DR: A novel approach to broadband log‐periodic antenna design is presented, where some of the most powerful evolutionary algorithms are applied and compared for the optimal design of wire log‐ periodic dipole arrays (LPDA) using Numerical Electromagnetics Code.
Abstract: Broadband log-periodic antenna optimization is a very challenging problem for antenna design. However, up to now, the universal method for log-periodic antenna design is Carrel's method dating from the 1960s. This paper compares five antenna design optimization algorithms (Differential Evolution, Particle Swarm, Taguchi, Invasive Weed, Adaptive Invasive Weed) as solutions to the broadband antenna design problem. The algorithms compared are evolutionary algorithms which use mechanisms inspired by biological evolution, such as reproduction, mutation, recombination, and selection. The focus of the comparison is given to the algorithm with the best results, nevertheless, it becomes obvious that the algorithm which produces the best fitness values (Invasive Weed Optimization) requires very substantial computational resources due to its random search nature. Log-periodic antennas (LPDA: Log-Periodic Dipole Arrays) are frequently preferred for broadband applications due to their very good directivity characteristics and flat gain curve. The purpose of this study is, in the first place, the accurate modelling of the log-periodic type of antennas, the detailed calculation of the important characteristics of the antennas under test (gain, vswr, front-to-back ratio) and the confrontation with accurate measurements results. In the second place, various evolutionary optimization algorithms are used, and notably the relatively new (2006) Invasive Weed Optimization (IWO) algorithm of Mehrabian & Lucas, for optimizing the performance of a log-periodic antenna with respect to maximum gain, Side-Lobe-Levels (SLL) and matching to 50 Ohms, VSWR. The multi-objective optimization algorithm is minimising a so-called fitness function including all the above requirements and leads to the optimum dipole lengths, spacing between the dipoles, and dipole wire diameters. In some optimization cases, a constant dipole wire radius is used in order to simplify the construction of the antenna. Fig. 1 is depicting the main antenna geometrical characteristics.

17 citations


Book ChapterDOI
13 Nov 2015
TL;DR: The traditional data management methods are not suitable for High-Performance Computing (HPC) systems and the files in such systems do not have semantic annotations and cannot be archived and managed for public dissemination.
Abstract: Scientific research activities generate a large amount of data, which varies in format, volume, structure and ownership. Although there are revision control systems and databases developed for data archiving, the traditional data management methods are not suitable for High-Performance Computing (HPC) systems. The files in such systems do not have semantic annotations and cannot be archived and managed for public dissemination.

3 citations


Book ChapterDOI
14 Feb 2015
TL;DR: This paper presents the methods for parallelising two pieces of scientific software, leveraging multiple GPUs to achieve up to thirty times speed up.
Abstract: The rate of scientific discovery depends on the speed at which accurate results and analysis can be obtained. The use of parallel co-processors such as Graphical Processing Units (GPUs) is becoming more and more important in meeting this demand as improvements in serial data processing speed become increasingly difficult to sustain. However, parallel data processing requires more complex programming compared to serial processing. Here we present our methods for parallelising two pieces of scientific software, leveraging multiple GPUs to achieve up to thirty times speed up.

1 citations


Proceedings ArticleDOI
28 Jul 2015
TL;DR: This work focuses on High Throughput Condor HTCondor as one of the most popular middlewares among UK universities, and proposes a new system PBStoCondor, capable of interfacing with Linux based system within the campus grids, and automatically determining the best resource for a given job.
Abstract: The campus grid architectures currently available are considered to be overly complex. We have focused on High Throughput Condor HTCondor as one of the most popular middlewares among UK universities, and are proposing a new system for unifying campus grid resources. This new system PBStoCondor is capable of interfacing with Linux based system within the campus grids, and automatically determining the best resource for a given job. The system does not require additional efforts from users and administrators of the campus grid resources. We have compared the real usage data and PBStoCondor system simulation data. The results show a close match. The proposed system will enable better utilization of campus grid resources, and will not require modification in users' workflows.

1 citations


Proceedings ArticleDOI
11 Dec 2015
TL;DR: IMTeract was used for energy usage profiling of HPC clusters running FLUENT and DL-POLY software and a GPU cluster running different implementations of an FFT algorithm, and the experimental results are encouraging and it is suggested that the IMTeract tool can be used to measure the CPU, Memory, Disk I/O and NetworkI/O for an application or a process and report on the energy used.
Abstract: Energy usage of computing equipment is an important consideration and energy inefficiency of computer systems is identified as the single biggest obstacle to advances in computing. Research into low-energy computing products ranges from operating system codes, applications and energy-aware schedulers to cooling systems for data centres. To monitor energy consumption in data and HPC centres it is necessary to develop tools for measuring the energy usage of computer equipment and applications. We have developed power measuring apparatus and a tool, IMTeract, for measuring energy consumption of HPC applications. IMTeract was used for energy usage profiling of HPC clusters running FLUENT and DL-POLY software and a GPU cluster running different implementations of an FFT algorithm. Our experimental results are encouraging and suggest that the IMTeract tool can be used to measure the CPU, Memory, Disk I/O and Network I/O for an application or a process and report on the energy used.