D-grid
The D-Grid Initiative is a grid computing initiative in the Federal Republic of Germany .
D-Grid aims to build a sustainable grid infrastructure for research and development in both academic and industrial areas in Germany. The development phase has been funded since September 1, 2005 by the Federal Ministry of Education and Research (BMBF) through several rounds of tenders with more than 70 million euros . As part of this funding, more than 20 projects were initiated in which more than 100 German research institutions participate. Furthermore, computing and storage infrastructure was made available through special measures in order to guarantee a smooth transition to grid operation without impairing the current infrastructure.
The D-Grid projects can be divided into basic services , higher grid services , academic disciplines and commercial users .
Basic services
Basic services are understood as a range of services that are indispensable for the functionality of a grid. This includes the provision of grid middleware and the establishment of an operating concept that covers security aspects and the formation of virtual organizations (VO). These basic services are provided by the D-Grid integration project (DGI 1: Sept. 2005 - Aug. 2007; DGI 2: Jan. 2008 - Dec. 2010). In order to react to new requirements from users, so-called gap projects were set up, which develop special basic services and work closely with the integration project.
Higher services
Higher grid services relate to requirements that are not generally necessary for the operation of a grid, but cannot only be used by one discipline. Examples of these services are services for resource management, for setting up and monitoring service level agreements (SLA) or for data storage. Projects developing these services ensure that the services build on the basic services and can be easily integrated by different disciplines.
Academic disciplines
Many scientific disciplines use large amounts of data that are made available to all scientists in a discipline. Due to the spatial distribution of the scientists, the use of a grid structure is inevitable. Projects in the fields of astronomy , high energy physics , climate research , medicine , engineering and the humanities have been set up to promote the use of the grid . Within these projects, discipline-specific grid services are developed. The respective projects are intended to lay the foundation for discipline-oriented grid structures that, if possible, support all scientists in the respective discipline in using the grid.
Commercial users
While large companies with several locations, often spread around the world, often set up their own enterprise grid , this is often not possible for smaller companies. Although these companies increasingly need access to IT resources, the acquisition and operation of these resources are associated with high costs for them, especially since the demand occurs in batches and not continuously. The grid with independent service and resource providers is a promising alternative for such smaller companies if the services are tailored to the needs of the company. Such services are to be developed in cooperation with companies in the corresponding projects.
Project examples
Some of the D-Grid initiative's projects are described in more detail below:
AstroGrid-D
In AstroGrid-D (also German Astronomy Community Grid, GACG ) a total of 13 scientific institutions from the fields of astronomy and computer science as well as high-performance computing centers work together. The main goal is to integrate the astronomical research institutes in Germany into a uniform grid-based infrastructure in order to promote distributed, collaborative work. Existing hardware and software resources, etc. a. astronomical data archives and robotic telescopes in the participating research institutes are to be integrated.
AstroGrid-D supports the standards of the International Virtual Observatory Alliance (IVOA) and works together with international grid projects. The AstroGrid-D is managed by the Astrophysical Institute Potsdam (AIP).
C3 grid
In the Collaborative Climate Community Data and Processing Grid , or C3-Grid for short , there is a network of climate researchers . This project deals not only with climate research, but also with the interactions of the climate system with socio-economic systems. In climate research, complex models are used to simulate the behavior of the earth system . The resulting model data is collected in archives at various locations in addition to the continuously accumulating observation data from the weather services. For a comprehensive (global and regional) analysis of this data, it is necessary to network these archives with one another and process them in the C3 grid. Due to the evaluation and storage of satellite data, there is a steadily growing need for storage space and computing capacity in climate research. The possibility of storing and analyzing the resulting data in a grid is intended to support the work of the climate researchers and their collaboration. The C3-Grid is managed by the Alfred Wegener Institute for Polar and Marine Research (AWI) in Bremerhaven .
D-Grid IaaS
At the beginning of grid computing, proprietary implementations and the like were used for the various grid middlewares. a. developed in the areas of authorization, monitoring and scheduling. The trend towards standardization and alignment of grid middleware has only existed for a few years. Cloud computing is experiencing a similar, but accelerated, development. Cloud computing offers can be IaaS, but also SaaS or PaaS. However, the term cloud computing is often only associated with Amazon's EC2 offer.
Until now, D-Grid resources have been accessed via the three grid middlewares gLite, Globus Toolkit and UNICORE that have been established in D-Grid. These middlewares were originally developed for an academic clientele and are largely non-intuitive to use. Particularly when attracting small and medium-sized companies as new customers, the associated complexity is a deterrent and is reduced by the EC2 offer. The fourth middleware pillar of the D-Grid is to be filled out by a Compute Cloud Middleware, the eCloud Manager. The current version already supports the control of VMware, Xen and Hyper-V, which means that D-Grid resource providers remain free to choose their virtualization environment.
The Institute for Robot Research at the Technical University of Dortmund and fluid Operations AG are involved in the D-Grid IaaS project .
GDI grid
The geodata infrastructures grid forms a project in a consortium of universities and companies to introduce existing geodata infrastructures to grid technologies. The sometimes very computationally intensive processes and algorithms that geoscientists use for modeling and simulation should benefit from the national grid infrastructure by using computing and storage capacities in a decentralized manner. First of all, frequently repetitive processing tasks (such as tessellation ) should be outsourced to grid processes. Furthermore, the very high volume of data in spatial data infrastructures (e.g. DTM , map and metadata catalogs) should be made as decentralized as possible and accessible to all authorized persons within the framework of the applicable license provisions.
The linking of different processing steps to clear application models is demonstrated by three example scenarios:
- In a flood simulation scenario, disasters like the Elbe flood are simulated by combining grid and GDI technologies.
- The spread of traffic noise in a city's road network and its impact on the surrounding residential and commercial areas is the subject of a second simulation.
- A third scenario is to provide rescue workers with the option of creating dynamic route planning in the event of a disaster, with the means of traditional vehicle navigation being expanded to include grid-based calculation and the consideration of obstacles (e.g. from fallen trees or rubble).
The GDI-Grid project is led by the Regional Computing Center for Lower Saxony ( RRZN ) at the University of Hanover .
HEP grid
The HEP Community Grid , or HEP Grid for short , is the German high-energy physics grid. This project is managed by the German Electron Synchrotron (DESY) in Hamburg , nine German institutes and universities as well as a number of associated partners are involved.
The main objective of the HEP grid is to improve data evaluation in high energy physics through the efficient use of distributed and networked storage and computer resources. The planned developments are important additions to the grid software used from the projects Enabling Grids for E-science (EGEE) and LHC Computing Grid (LCG). They are an important contribution to the data analysis of current and future large-scale experiments, for example at the Large Hadron Collider (LHC) at CERN or the planned International Linear Collider (ILC).
InGrid
The InGrid community project is an association in the field of engineering. InGrid enables grid-based applications and the efficient use of shared computer and software resources for engineering projects. As part of this project, a grid environment for engineering applications is to be created. Through the flexible use of grid technologies, modeling, simulation and optimization skills are to be brought together and the shared use of resources is to be made possible efficiently.
Five typical application areas (foundry technology, forming, groundwater flow and transport, turbine simulation and interaction of flow and structural mechanics) are to be worked on as examples in order to cover the three central areas of computationally intensive engineering applications (coupled multiscale problems, coupled multidisciplinary problems and distributed simulation-based optimization). In particular, adaptive and scalable process models and grid-based process environments are developed for these tasks.
By its very nature, engineering research is application and industry-oriented. The support of virtual prototyping and the optimization of engineering processes is therefore a focus of the project. InGrid is managed by the High Performance Computing Center (HLRS) at the University of Stuttgart .
MediGRID
In the joint project MediGRID , renowned research institutions in the fields of medicine, bioinformatics and health sciences have come together as consortium partners to develop a grid middleware integration platform and eScience services for biomedical science based on it. The further involvement of numerous associated partners from industry, supply and research institutions puts the project on a broad basis of interest.
A modular distribution of tasks was chosen for the project: In the four methodical project modules ( middleware, ontology tools, resource fusion and eScience ) the relevant consortium partners are gradually developing a grid infrastructure. In particular, they take into account the requirements of grid users from the biomedical environment, which are exemplarily developed in the three application-oriented project modules ( biomedical informatics, image processing, clinical research ).
SuGI
SuGI - Sustainable Grid Infrastructure - is a gap project of the D-Grid initiative with the aim of bringing grid computing to the surface and making it usable there. The project is geared towards the large number of small and medium-sized data centers at universities and companies that have not yet used grid technologies or only use them to a limited extent. You will be supported in the provision of grid resources and services.
By setting up a scalable training infrastructure (SuGI training portal), providing technical assistance with the installation and operation of the middleware, and developing and evaluating legal and organizational structures, SuGI makes a sustainable contribution to the creation of a grid platform for e-science in Germany.
TextGrid
Although e-science concepts and grid technologies originally come from the natural sciences and medicine , there are also broad areas of application in the humanities and art . TextGrid is the first humanities grid project in Germany and has therefore been involved in the "gridification" of the humanities together with other e-Humanities initiatives from the very beginning.
In this way, TextGrid makes a contribution to text-based research. A grid-compatible workbench enables the collaborative philological processing, analysis, annotation, edition and publication of scientific texts. Grid technologies connect the existing text archives and - inspired by approaches and development opportunities for the Semantic Grid - link their resources such as text corpora, dictionaries and verification instruments with one another. The interfaces, which are open for further projects, enable synergies with other initiatives in scientific word data processing as well as a rationalization of scientific work, among other things through optimized access to primary sources and tools.
The following specialist and technical partners cooperate in TextGrid: the SUB Göttingen (project management), the Institute for German Language , the Max Planck Digital Library , the University of Trier , the University of Worms , the University of Paderborn , the Technical University of Kaiserslautern , the Ludwig-Maximilians- University of Munich , the University of Würzburg and DAASI International GmbH.
WISENT
WISENT is the knowledge network for energy meteorology - an e-science project funded by the BMBF for information technology-optimized collaboration between organizations that conduct research and development in the field of energy meteorology. The focus is on renewable energies , which are particularly dependent on the weather.
The partners in WISENT are the German Aerospace Center (DLR), the University of Oldenburg , the Oldenburg Research and Development Institute for IT Tools and Systems (OFFIS) and meteocontrol GmbH .
With the construction of the virtual Institute for Energy Meteorology (vIEM) funded by the Helmholtz Association , an institutional basis has been created which, in the IT area, now needs to be supplemented by innovative grid-based technologies for access to distributed and often very heterogeneous resources and their distributed processing power. The characteristic of the work in the vIEM is the joint use and processing of large amounts of data (in the order of magnitude of many terabytes), for which sufficiently powerful instruments and services are not yet available. The parallel processing of the data represents a promising solution to these problems.
ValueGrids
ValueGrids designs and develops an integrated approach for service level management in value networks . This enables providers of Software-as-a-Service to use the D-Grid infrastructure sustainably.
The cooperation partners of ValueGrids are: SAP AG (coordination), Conemis AG , IBM Deutschland Research & Development GmbH , Albert-Ludwigs-Universität Freiburg and the Karlsruhe Institute of Technology .
Web links
Individual evidence
- ↑ SuGI training portal ( Memento of the original from March 5, 2009 in the Internet Archive ) Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice.
