Catálogo de publicaciones - libros

Between high-performance clusters and grids appears an intermediate infrastructure called cluster grid that corresponds to the interconnection of clusters through the Internet. Cluster grids are not only dedicated to specific applications but should allow the users to execute programs of different natures. This kind of architecture also imposes additional constraints as the geographic extension raises availability and security issues. In this context, authentication is one of the key stone by providing access to the resources. Grid5000 is a french project based on a cluster grid topology. This article expounds and justifies the authentication system used in Grid5000. We first show the limits of classical approaches that are local files and NIS in such configurations. We then propose a scalable alternative based on the LDAP protocol allowing to meet the needs of cluster grids, either in terms of availability, security and performances. Finally, among the various applications that can be executed in the Grid5000 platform, we present grid, a minimal middleware used for medical data processing.

Compute-intensive simulations are currently good candidates for being executed on distributed computers and Grids, in particular for applications with a large number of input data whose values change throughout the simulation time and where the communications are not a critical factor. Although the number of computations usually depends on the bulk of input data, there are applications in which the computational load depends on the particular values of some input data. We propose a general methodology to deal with the problem of improving load balance in these cases. It is divided into two main stages. The first one is an exhaustive study of the parallel code structure, using performance tools, with the aim of establishing a relationship between the values of the input data and the computational effort. The next stage uses this information and provides a mechanism to distribute the load of any particular simulating situation among the computational nodes. A load balancing strategy for the particular case of STEM-II, a compute-intensive application that simulates the behavior of pollutant factors in the air, has been developed, obtaining an important improvement in execution time.

Autonomic and Grid computing are complementary, in the sense that complex grid environments can take advantage of the features provided by autonomic computing and on the other hand, autonomic processes can be properly deployed by using grid technology. Besides, one of the most active fields in grid computing is the area of data grids, focusing on the data access. Our paper proposes a grid framework which provides autonomic characteristics in order to enhance the performance of the data access, predicting the future behaviour of the corresponding I/O system. The use of the autonomic system is transparent to the user. This paper also presents a study case of such system.

TCP is the only protocol widely available for reliable end-to-end congestion-controlled network communication, and thus it is the one used for almost all communications. Unfortunately, TCP is not designed with high-performance networking and computing. Thus the research for TCP to obtain good throughput in high-performance networking and computing is in progress all over the world actively. In this paper, we propose a new scheme which makes a TCP system achieve high throughput even with small buffer. The receive buffer almost empties due to the characteristic of original TCP but the amount of physical memory assigned for the buffer cannot be reduced because TCP flow control will downgrade TCP performance with the reduced buffer. However a TCP system applying our proposed scheme can reduce the size of physically assigned receive buffer without downgrading TCP performance. And then we use PID control mechanism as a tool to adjust the size of VRB properly. Lastly, we compare the throughput with two schemes, proposed scheme and original TCP scheme. As a result, the TCP using VRB obtains 46% higher throughput than the original one. And we also compare the amount of memory necessary for achieving the maximum throughput between two schemes. The result of second comparison shows that the proposed TCP spends 43% less memory than the tuned original TCP for same throughput.

Development of a coarse-grain parallel algorithm of artificial neural networks training with dynamic mapping onto processors of parallel computer system is considered in this paper. Parallelization of this algorithm done on the computational grid operated under Globus middleware is compared with the results obtained on the parallel computer Origin 300. Experiments show better efficiency for computational grid instead of parallel computer with an efficiency/price criterion.

The paper presents a solution of the problem of wrapping relational databases to an object-oriented business model in the data grid architecture. The main problem with this kind of wrappers is how to utilize the native SQL query optimizer, which in majority of RDBMS is transparent for the users. In our solution we use the stack-based approach to query languages, its query language SBQL, updateable object-oriented virtual views and the query modification technique. The architecture rewrites the front-end OO query to a semantically equivalent back-end query addressing the M0 object model that is 1:1 compatible with the relational model. Then, in the resulting SBQL query the wrapper looks for patterns that correspond to optimizable SQL queries. Such patterns are then substituted by dynamic SQL statements. The method is illustrated by a sufficiently sophisticated example. The method is currently being implemented within the prototype OO server ODRA devoted to Web and grid applications.

This paper proposes a method for modeling views in Grid databases. Views are understood as independent data transformation services that may be integrated with other Database Grid Services. We show examples of graphical notation and semi-automated query construction. The solution is supported by a minimal metamodel, which also provides reflection capabilities for dynamic services orchestration.

Caching can highly improve performance of query processing in distributed databases. In this paper we show how this technique can be used in grid architecture where data integration is implemented by means of updatable views. Views integrate data from heterogeneous sources and provide users with their integrated form. The whole process of integration is transparent, i.e. users need not be aware that data are not located at one place. In data grids caching can be used at different levels of architecture. We focus on caching at the middleware layer where the cache is stored in the database of the integrating unit. These results can be then used while answering queries from grid users, so there will be no need to reevaluate the whole queries. In such a way caching can highly increase performance of applications operating on grid. In the paper we also present an example how a query can be optimized by rewriting to make use of cached results.

This paper presents the + ROWA model that is been develop at the Universidad Politecnica de Madrid with the aim of replicating information in Grid environments and optimizing the number of messages to be exchanged in the process as well as their use to built Grid environments based on WSRF specifications. The model presented in this paper will be one of the pillars of a new Grid service (WS-ReplicationResource) that, in the near future, will provide Grid systems with a high level service of resource replication.

Modern enterprises face a strong economical pressure to increase competitiveness, to operate on a global market, and to engage in alliances of several kinds. Agility has become the new guiding principle for enterprises. It means that enterprises must be able to easily collaborate with other enterprises expand, be it through participating in enterprise networks, or through mergers or acquisitions, or through insourcing or outsourcing of services. In order to meet these economical requirements, enterprises rely increasingly on the benefits of modern information and communication technology (ICT). However, the appropriate knowledge to deploy this technology as needed, and in an effective and efficient way, is largely lacking, particularly knowledge regarding the collaboration of enterprises and knowledge regarding the interoperability of their information systems.