A Grid-based architecture for Multimedia services management Dario Bruneo*, Mirko Guarnera**, Angelo Zaia*, Antonio Puliafito* (*)Dept. of Mathematics, University of Messina Salita Sperone - Contrada Papardo, 98166 Messina, Italy (**)STMicroelectronic, AST (Advanced System Technology)- Catania Lab, Stradale primosole 50, 95121 Catania, Italy {dbruneo,apulia,azaia}@ingegneria.unime.it
[email protected]
The current client server computational model assumes to know where computing and storage power are located. As imposed from the current technological evolution, as well as from the continuously increasing users requirements, it is widely recognized that such a model has to evolve in order to transparently access distributed computing and storage resources, showing the final results directly on the user terminal. Thus it is mandatory to develop a middleware layer that will mediate and manage the access to the distributed resources. Grid technology [7] [6] is a new paradigm which has the potential to completely change the way of computing and data access. Although there is no widely agreed definition what Grid means, the consequences for scientific and research work, but also for the private users, will be severe. Computational Grids are widely regarded as the next logical step in computing infrastructure, following a path from standalone systems, to tightly linked clusters, to enterprise-wide clusters, to geographically dispersed computing environments. Generally speaking, we could consider the Grid as the new enabling technology to transparently access computing and storage resources anywhere, anytime and with guaranteed Quality of Service (QoS). Although still in its infancy, Grid is already being successfully used in many scientific applications where huges amounts of data have to be processed and/or stored. Some examples include the processing of data coming from physic-related applications, like the data generated in the nuclear accelerator of CERN [1], the data coming from radio telescopes, the data generated in complex simulations [2] and so on. There is no doubt that such demanding applications have created, justified and diffused the concept of Grid among the scientific community, but as a matter of fact such powerful technology is still confined to a very restricted scientific environment. We do think that the time is right for the Grid to reveal its potentialities to a wider audience, which is mainly constituted of users with every day problems and requirements. Obviously such needs cannot directly compete with the highly demanding scientific applications described so far, but, as the amount of potential users is really enormous, the accumulated data processing and storage requirements are at least comparable. One of the possible new application fields for the Grid Computing is the management of multimedia services in wireless environments. In fact the current tendency is to allow mobile users to access
powerful services as Video On Demand. Mobility poses new and challenging issues to be solved, among which the QoS management is one of the most crucial, due to the immediate consequences it has on the users satisfaction. Our idea is to use the Grid paradigm to hide both implementation and computational details, as sophisticated middleware layers take care of resource management, load balancing, security issues, and resources optimization in general. The reference scenario is showed in Fig. 1.
Fig. 1. The reference scenario.
In [4] we have proposed a reference architecture that makes use of Grid computing for QoS management of MPEG-4 flows. In this paper we focus on the Grid zone and we present an agent-based architecture for the fulfillment of a Grid node. We use the Grid for a transparent and effective management of the computational resources and of the data present in the Wired Area. In fact, the Grid allows to move data, so that it can be made ”closer” to the Access Points where the user is. The tailoring operations will be easier by moving some parts of code. Furthermore, some load balancing operations will be possible by assigning some tasks to the computational resources present in the Grid zone. The performed task by each Grid node shall be: – network workload balance based on the users’ displacement – speeding up of tailoring operations on movies, by the distribution of the operations themselves on a lot of nodes In order to perform these tasks, several operations shall be necessary. Some of them shall be at a low level, such as the computational resources management, which are present on each node, and the remote jobs submission. Other operations shall be at an high level, such as the discovery of connected nodes
and the management of the users’ information (mobility, terminal type, etc). In order to solve the previous issues, this paper proposes, for the Grid Node, an agent-based architecture as shown in Fig. 2. As we can see in Fig. 2, each node of the Grid network will consist of a two-level architecture. The lowest one is the level that provides the Grid basic services (resource management, security, distributed access).
Fig. 2. Grid Node.
An agent platform containing the Grid Agents will be present at the highest level of the architecture. The task of the Grid Agents will be to respond to the queries coming from the other agents present in the middleware in order to satisfy the requests. Moreover, the Grid Agents will have the task to manage the issues related to Discovery of Services, network load balancing and to the availability of the resources. In order to achieve the Discovery of Services we plan to use a peer to peer [5] approach by integrating JXTA technology [3][8] inside the Grid nodes. Thus each node will represent a peer which will be able to discover other peers and the services offered.
References 1. Cern openlab for datagrid applications. http://proj-openlab-datagridpublic.web.cern.ch/proj-openlab-datagrid-public/. 2. The datagrid project. http://www.datagrid.cnr.it/. 3. JXTA web site. http://www.jxta.org/.
4. D. Bruneo, M. Villari, A. Zaia, and A. Puliafito. Qos management for mpeg-4 flows in wireless environment. to be published in Microprocessors and Microsystems, Elsevier Science. 5. D. Clark. Face-to-face with peer to peer networking. Computer, 34(1):18–21, Jan. 2001. 6. D. W. Erwin and D. F. Snelling. UNICORE: A Grid computing environment. Lecture Notes in Computer Science, 2150, 2001. 7. I. Foster. The anatomy of the Grid: Enabling scalable virtual organizations. Lecture Notes in Computer Science, 2150, 2001. 8. l. Gong. JXTA: A network programming environment. IEEE Internet Computing, 5(3):88–95, May/June 2001.