| Data Management in Grid Middleware | |
| Researcher | Jon Kristian Nilsen, University of Oslo |
| Supervisors |
Prof. Alexander Read, Dept. of Physics (UiO) Dr. Jostein K. Sundet, Manager, Scientific Computing Group (UiO) |
| Description | |
| Status | Initial study |
| Documents, links | |
| Management of Scientific Data in Grid Systems | |
| Researcher | Salman Zubair Toor, Uppsala University |
| Supervisors |
Dr. Sverker Holmgren, IT Institute (UU) Dr. Mattias Ellert, Nuclear and Particle Physics Institute (UU) |
| Description | The project concentrates on the Storage Resource Manager standard and aims to enhance/modify it in a way that it will provide more facilities to the ARC users. We are aiming to initiate a research study plan of the whole architecture and try to build some of the components that can work on top of the separate SRMs. This new layer will again publish interfaces for the client applications as SRM is currently doing but this time these interfaces are not SM dependent. For this new layer we will follow the standard way of publishing interfaces by using web services (maybe we will use the GT4 development environment). The main idea behind this research plan is eventually to study the possibilities of making the different SMs transparent for the ARC clients and to create an environment for the site administrators in which they can single handedly handle the different SMs. The proposed layer will not be only a message passing interface between client and the SRMs. In fact, it will provide a complete cost based estimation system to find the best possible SM amongst the SM registered with this layer. This decision will be based on the information provided through the SRMs of those SMs. Since this layer is on top of the SRMs thus through this we can also replicate the data not only within a single SM but also within the different SMs under the GSRMs. This will greatly reduce the dependency of any site on a single storage manager. |
| Status | Initial study |
| Documents, links | |
| Distributed Data Analysis with AliEn and ARC | |
| Researcher | Philippe Gros, Lund University |
| Supervisors |
Prof. Hans-Åke Gustafsson, Div. for Experimental High Energy Physics (LU) Dr. Peter Christiansen, Div. for Experimental High Energy Physics (LU) Dr. Oxana Smirnova, Div. for Experimental High Energy Physics (LU) |
| Description | |
| Status | Initial study |
| Documents, links | |
| Fine-grained delegation in Grids | |
| Researcher | Mika Silander, Helsinki Institute of Physics |
| Supervisors | Miika Tuisku (HIP) |
| Description |
Current activity: Modeling the way fine-grained rights delegations are expressed. The most promising standard for this purpose is SAML. There are however, several aspects remaining that directly or indirectly affect the genericity and expressiveness of rights delegations. Recent efforts have focused on identifying and analysing their implications. Plans: Discuss the open issues regarding the level of genericity with members of the NorduGrid community in order to find a representative set of delegation use cases that fulfills the needs of ARC users. Based on feedback and use cases, define a format for rights expressions. The definition itself is likely to be expressed as an XML schema. Once the definition is ready, the focus will turn towards implementing a solution that would be compatible with both message level and transport level security mechanisms. The solution itself is to conform with the architecture described in the KnowARC Design document. |
| Status | Initial study |
| Documents, links | |
| Automatic Job Resubmission | |
| Researchers | Henrik Jensen, Jesper Leth (Aalborg University) |
| Description | A daemon which supervises jobs and resubmits them if they fail. |
| Type | Master project |
| Status | Completed (January 2004) |
| Links |
Repository Report |
| VCRRecord | ||
| Researchers | Henrik Høy Karlsen (SDU) | |
| Description | A Grid VCR (video recorder) using a TV tuner card and capture scripts based on Xawtv. | |
| Type | Course assignment | |
| Status | Completed (February 2004) | |
| Links | Description and examples are available at http://www.imada.sdu.dk/~karlsen/vcrrecord.html | |
| Survey of Grid banking tools | |
| Researchers | Kristian Frogner, Stein E. Wethal, Tarjei Mandt (HIG) |
| Description | The main goal is to survey existing grid banking tools and assess their applicability to NorduGrid*. This also includes integration efforts of existing systems into the NorduGrid software or a proposal for NorduGrid's own accounting system. |
| Type | Bachelor degree project |
| Status | Completed (May 2004) |
| Links | Report |
| SWEGRID accounting system | |
| Researchers | Erik Elmroth, Martin Folkman, Peter Gardfjäll, Olle Mulmo, Tomas Olsson, Åke Sandgren, Thomas Sandholm (PDC) |
| Description | Developing a coordinated Accounting System for the SweGrid clusters |
| Type | R&D project |
| Status | In progress |
| Links | SWEGRID SGAS homepage |
| Condor backend | |
| Researchers | Haakon Riiser (UiO) |
| Description | Development of Condor backend for the NorduGrid middleware. |
| Type | Diploma project |
| Status | In progress |
| Links | |
| Benchmark-based scheduling | |
| Researchers | Johan Tordsson, Erik Elmroth (UmU) |
| Description | Development of algorithms, methods, and software for a grid resource manager, responsible for resource brokering and scheduling in early production grids. The broker selects computing resources based on actual job requirements and a number of criteria identifying the available resources, with the aim to minimize the total time to delivery for the individual application. The total time to delivery includes the time for program execution, batch queue waiting, input/output data transfer, executable staging, etc. Main features include support for making advance reservations, to make resource selections based on computer benchmark results and network performance predictions, and to enable a basic adaptation facility. |
| Type | Doctoral study |
| Status | In progress |
| Links |
PARA04 talk Report |
| GRID Modelica | |
| Researchers | Kaj Nyström (LiU) |
| Description | The aim is to combine the power of grid computing with a modern high level object oriented modelling language in order to create a platform for modelling on the grid, primarily for industrial applications. The language to work on is the Modelica language and the compiler used in the development is the OpenModelica compiler. The research will be conducted in conjunction with industrial partners such as SKF and MathCore Engineering. |
| Type | Doctoral study |
| Status | Starting |
| Links | Modelica OpenModelica compiler |