|
Grid Middleware course (2004)
Content (3/4):
Course 9: The Globus Toolkit Data management Course 10: The Globus Toolkit security infrastructure Course 11: VLAM-G: Virtual laboratory for experimental science (E.C Kaletas/V. Korkhov, UvA) Course 12: Ibis: A Java-based grid programming environment (Henri Bal, VU) Course 13: European Data Grid Project (David Groep, NIKHEF) Course 14: How to Dynamically Provision a Network Wavelength. (C.M.T Delaat, UvA ) Introduction to Java (optional course) Assignments Other stuff
Course 9: GT3 Data ManagementDate: 19-01-2004 Description: This course describes Globus toolkit components dealing with data manipulation such as the Reliable Transfer Service (RFT), the Gridftp protocol, the data replication service, or the Globus component to access to secondary storage. Keywords: Course material: the slides [PPT] References:
Course 10: GT3 Security ManagementDate: XX-XX-XXXX Description: This course Has been cancelled The slides are given for the students interested in this topic. The course is discussing security issues and how they have been dealt with in the Globus Project. The Globus security infrastructure is designed to bring a solution to three challenges issues: integration, interoperability, and trust in VO. Integration solutions where existing services needs to be used, and interfaces should be abstracted to provide an extensible architecture. Interoperability solutions so that services hosted in different virtual organizations that have different security mechanisms and policies will be able to invoke each other. And finally, solutions to define, manage and enforce trust policies within a dynamic Grid environment. Keywords: Course material: the slides [PPT] References:
Course 11: Introduction to Java BasicsDate: 15-01-2004 Description: Keywords: Course material: the slides [PPT] (Examples) References:
Course 12:Date: 22-01-2004 Description: VLAM-G, the Grid-based Virtual Laboratory AMsterdam, provides a science portal for distributed analysis in applied scientific research. Science portals generally provide easy access to problem solving environment. VLAM-G science portals offer to scientists with a limited knowledge in computer technology the possibility to make efficient use of the computation power, experimentation resources and the network infrastructure available to them without having to know all the technical details. In this lecture we are going to review the design and the implementation issues of a major component of the VLAM-G architecture the VLAM-G information system (VIMCO: Objectives, Functionality Design & Implementation, and Future work). In second half of the course, we will have an overview of resource Management in the context of the VLAM-G/VL-e project. This talk presents a review of resource management issues in Grid environments and the state of art in Grid resource management. The invention of meta-computing and introduction of Grid technologies to support distributed computations and data processing led to the necessity of significant changes in resource management strategies for distributed heterogeneous systems. These changes reflect the dynamic, shared and heterogeneous nature of Grid environment aiming at the provision of effective meta-scheduling mechanism for the new computing environment. The design and implementation issues of resource management component of VLAM-G, the Grid-based VLAM-G/VL-e, are also going to be reviewed in this lecture. Keywords: Course material: the slides [PPT,PDF] References: The Web page of the VL-e project
Course 13: Ibis A Java-based grid programming environmentDate: 26-01-2004 Guest Lecturer: Pro. Henri Bal, VU Description: Ibis is an ongoing research project in which we are building a Java-based Grid programming environment for distributed supercomputing applications. Java's high portability allows parallel applications to run on a heterogeneous grid without requiring porting or recompilation. A major problem in using Java for high-performance computing, however, is the inferior performance and limited expressiveness of Java's Remote Method Invocation (RMI). Earlier projects (e.g., Manta) solved the performance problem, but at the cost of using a runtime system written in native code, which gives up Java's high portability. The philosophy behind Ibis is to try to obtain good performance without using any native code, but allow native solutions as special-case optimizations. For example, a Grid application developed with Ibis can use a pure-Java RMI implementation over TCP/IP that will run "everywhere"; if the application runs on, say, a Myrinet cluster, Ibis can load a more efficient RMI implementation for Myrinet that partially uses native code. The pure-Java implementation of Ibis does several optimizations, using bytecode rewriting. For example, it boosts RMI performance by avoiding the high overhead of runtime type inspection that current RMI implementations have. The Ibis programming environment consists of a communication runtime system with a well-defined interface and a range of communication paradigms implemented on top of this interface, including RMI, object replication, group communication, and collective communication, all integrated cleanly into Java. Ibis has also been used to implement Satin, which is a Cilk-like wide-area divide-and-conquer system in Java. Experiments have been performance on two Grid test beds, the Dutch DAS-2 system and the (highly heterogeneous) European GridLab test bed. Our current research on Ibis focuses on fault tolerance and on heterogeneous networks. Keywords: Course material: the slides [PPT] () References: Ibis: an efficient Java-based grid programming environment Prof. Henri Bal web page
Course 14: European Data Grid ProjectDate: 27-01-2004 Guest Lecturer: David Groep, NIKHEF Description:Many scientific applications today are entering into a new arena, where more detailed and better results can be obtained by collecting massive amounts of data. Whereas a few years ago datasets were typically in the order of megabytes, the next generation scientific problems will generate terabytes, if not petabytes of data. In all cases, either the scientists, the data acquisition, or both, are scattered around the world. In Europe, the DataGrid project was established in 2000 to address the problem of data-intensive computing in a distributed environment. In this lecture, we will give an overview of the EU DataGrid, its architecture, and the actual deployment of the largest Grid production facility in the world that is based on the EDG software suite: the LHC Computing Grid, that will involve over 1000 sites and 10 000 scientists from all over the world. By running through some concrete examples, some of the stronger points as well as the weakness of the currently deployed middleware will be illustrated. Keywords: Course material: the slides [PPT , PDF] References: |
|
Workflow & Grid Middleware Group (WGM) |
