Computational Technology Resources - CSETS

Keywords: DEISA, high performance computing, e-infrastructure, engineering applications, scientific applications.

Summary

This chapter briefly describes DEISA, the Distributed European Infrastructure for Supercomputing Applications, the vision, mission, and objectives, and recent achievements of the project with special emphasis on the areas of e-Infrastructures, infrastructure services, the grid middleware and its components, and user support and applications. The second part of the chapter deals with the DEISA Extreme Computing Initiative DECI, and presents the results of five grand-challenge engineering application simulations in the fields of wall turbulence, turbulent reacting flows, heavy particles in turbulent flows, noise generated by supersonic jets, and the flow over a bump.

In summary, as has been demonstrated, the DEISA team focuses strongly on e-Science end-users and their applications, to guarantee a user-friendly and collaborative high performance computing (HPC) environment, remote and secure access to all distributed resources, and seamless collaboration of (virtual) research teams. In addition, the DEISA team has learned that it is extremely important to avoid interference with well-established local policies of national HPC centers, and instead to concentrate on orchestrating the European HPC service as a layer on top addressing integration, interoperability, and standards.

Six years of DEISA production has shown that the concept implemented in DEISA has succeeded well. DEISA aimed at designing, deploying and operating a persistent European infrastructure for general purpose high performance computing. This turnkey operational HPC infrastructure is now gaining acceptance under the umbrella of PRACE (the Partnership for Advanced Computing in Europe) that is preparing for the integration of European Tier-0 and Tier-1 HPC services. As a central player for European High Performance Computing the DEISA Consortium is contributing to a global e-infrastructure for science and technology. The transition paths between national Tier-1 supercomputers and the new leadership Tier-0 petascale systems need to be as seamless as possible to open up the way to new research dimensions in Europe.

The operation of an infrastructure like DEISA has led to new management challenges not seen before. Managing a supercomputer system or a number of locally installed cluster systems differs strongly from a European supercomputer infrastructure where staff members dealing with the same problem are hundreds of miles away from each other. There is no short cut, going to the office next door, just checking if we agree on some option settings within a software component. Within a virtual organization every small modification has to be checked by all partners over and over again. Installing new software components requires synchronization by all participants, if any dependencies exist. Scheduling of tasks, installations, system power up and down, network infrastructure changes and others have to be agreed.

Our experience shows that many of these tasks cannot be handled only by e-mail. It is mandatory to have regular video or phone conferences, writing minutes and checking for completion of tasks. Additionally it is often necessary to agree on strict rules for processing, especially in the case of disagreements. Often, those dissents are found in security policy issues, scheduling of software installation and upgrades, and budget issues in the context of necessary components.

For this purpose the DEISA operation team has been established: planning and coordination of tasks, forwarding of information, power of decision, and managing in general are prerequisites for a successful European production quality infrastructure. Establishing this team has simplified greatly the collaborative work, and it should be recommended to anyone dealing with similar infrastructures that they start with adequate organizational and management structures.

purchase the full-text of this chapter (price £20)

go to the previous chapter
go to the next chapter
return to the table of contents
return to the book description
purchase this book (price £95 +P&P)

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Computational Science, Engineering & Technology Series ISSN 1759-3158 CSETS: 27 TRENDS IN PARALLEL, DISTRIBUTED, GRID AND CLOUD COMPUTING FOR ENGINEERING Edited by: P. Iványi, B.H.V. Topping Chapter 2 DEISA: A Distributed European HPC Ecosystem for Engineers and Scientists W. Gentzsch¹ and H. Lederer² ¹The DEISA Project and Open Grid Forum, Neutraubling, Germany ²Garching Computing Centre, Max Planck Society, IPP, Garching, Germany doi:10.4203/csets.27.2 purchase the full-text of this chapter Full Bibliographic Reference for this chapter W. Gentzsch, H. Lederer, "DEISA: A Distributed European HPC Ecosystem for Engineers and Scientists", in P. Iványi, B.H.V. Topping, (Editors), "Trends in Parallel, Distributed, Grid and Cloud Computing for Engineering", Saxe-Coburg Publications, Stirlingshire, UK, Chapter 2, pp 21-47, 2011. doi:10.4203/csets.27.2 Keywords: DEISA, high performance computing, e-infrastructure, engineering applications, scientific applications. Summary This chapter briefly describes DEISA, the Distributed European Infrastructure for Supercomputing Applications, the vision, mission, and objectives, and recent achievements of the project with special emphasis on the areas of e-Infrastructures, infrastructure services, the grid middleware and its components, and user support and applications. The second part of the chapter deals with the DEISA Extreme Computing Initiative DECI, and presents the results of five grand-challenge engineering application simulations in the fields of wall turbulence, turbulent reacting flows, heavy particles in turbulent flows, noise generated by supersonic jets, and the flow over a bump. In summary, as has been demonstrated, the DEISA team focuses strongly on e-Science end-users and their applications, to guarantee a user-friendly and collaborative high performance computing (HPC) environment, remote and secure access to all distributed resources, and seamless collaboration of (virtual) research teams. In addition, the DEISA team has learned that it is extremely important to avoid interference with well-established local policies of national HPC centers, and instead to concentrate on orchestrating the European HPC service as a layer on top addressing integration, interoperability, and standards. Six years of DEISA production has shown that the concept implemented in DEISA has succeeded well. DEISA aimed at designing, deploying and operating a persistent European infrastructure for general purpose high performance computing. This turnkey operational HPC infrastructure is now gaining acceptance under the umbrella of PRACE (the Partnership for Advanced Computing in Europe) that is preparing for the integration of European Tier-0 and Tier-1 HPC services. As a central player for European High Performance Computing the DEISA Consortium is contributing to a global e-infrastructure for science and technology. The transition paths between national Tier-1 supercomputers and the new leadership Tier-0 petascale systems need to be as seamless as possible to open up the way to new research dimensions in Europe. The operation of an infrastructure like DEISA has led to new management challenges not seen before. Managing a supercomputer system or a number of locally installed cluster systems differs strongly from a European supercomputer infrastructure where staff members dealing with the same problem are hundreds of miles away from each other. There is no short cut, going to the office next door, just checking if we agree on some option settings within a software component. Within a virtual organization every small modification has to be checked by all partners over and over again. Installing new software components requires synchronization by all participants, if any dependencies exist. Scheduling of tasks, installations, system power up and down, network infrastructure changes and others have to be agreed. Our experience shows that many of these tasks cannot be handled only by e-mail. It is mandatory to have regular video or phone conferences, writing minutes and checking for completion of tasks. Additionally it is often necessary to agree on strict rules for processing, especially in the case of disagreements. Often, those dissents are found in security policy issues, scheduling of software installation and upgrades, and budget issues in the context of necessary components. For this purpose the DEISA operation team has been established: planning and coordination of tasks, forwarding of information, power of decision, and managing in general are prerequisites for a successful European production quality infrastructure. Establishing this team has simplified greatly the collaborative work, and it should be recommended to anyone dealing with similar infrastructures that they start with adequate organizational and management structures. purchase the full-text of this chapter (price £20) go to the previous chapter go to the next chapter return to the table of contents return to the book description purchase this book (price £95 +P&P)
Back to top	©Civil-Comp Limited 2023 - terms & conditions