http://hdl.handle.net/2117/1666 Fri, 24 May 2013 21:47:06 GMT 2013-05-24T21:47:06Z webmaster.bupc@upc.edu Universitat Politècnica de Catalunya. Servei de Biblioteques i Documentació no http://hdl.handle.net/2117/2382 Title: The catallaxy approach for decentralized economic-based allocation in Grid resource and service markets. Authors: Ardáiz Villanueva, Óscar; Artigas Vidal, Pau; Eymann, Torsten; Freitag, Fèlix; Navarro Moldes, Leandro; Reinicke, Michael Abstract: Efficient resource allocation in dynamic large-scale environments is one of the cha llenges of Grids. In centralized economic-based allocation approaches, the user requests can be matched to the fastest, cheapest or most available resource. This approach, however, shows limitations in scalability and in dynamic environments. In this paper, we explore a decentralized economic approach for resource allocation in Grid markets based on the Catallaxy paradigm. Catallactic agents discover selling nodes in the resource and service Grid markets, and negotiate with each other maximizing their utility by following a strategy. By means of simulations, we evaluate the behavior of the approach, its resource allocation efficiency and its performance with different demand loads in a number of Grid density and dynamic environments. Our results indicate that while the decentralized economic approach based on Catallaxy applied to Grid markets shows similar efficiency to a centralized system, its decentralized operation provides greater advantages: scalability to demand and offer, and robustness in dynamic environments. Fri, 21 Nov 2008 10:34:40 GMT http://hdl.handle.net/2117/2382 2008-11-21T10:34:40Z Ardáiz Villanueva, Óscar; Artigas Vidal, Pau; Eymann, Torsten; Freitag, Fèlix; Navarro Moldes, Leandro; Reinicke, Michael no Catallaxy paradigm, Decentralized allocation, Dynamic large-scale environments, Economic-based allocation, Grid computing, Grid markets, Large scale systems, Resource allocation, Service grid markets, Utility programs Efficient resource allocation in dynamic large-scale environments is one of the cha llenges of Grids. In centralized economic-based allocation approaches, the user requests can be matched to the fastest, cheapest or most available resource. This approach, however, shows limitations in scalability and in dynamic environments. In this paper, we explore a decentralized economic approach for resource allocation in Grid markets based on the Catallaxy paradigm. Catallactic agents discover selling nodes in the resource and service Grid markets, and negotiate with each other maximizing their utility by following a strategy. By means of simulations, we evaluate the behavior of the approach, its resource allocation efficiency and its performance with different demand loads in a number of Grid density and dynamic environments. Our results indicate that while the decentralized economic approach based on Catallaxy applied to Grid markets shows similar efficiency to a centralized system, its decentralized operation provides greater advantages: scalability to demand and offer, and robustness in dynamic environments. http://hdl.handle.net/2117/2344 Title: Integration of decentralized economic models for resource self-management in application layer networks Authors: Chacín Martínez, Pablo; Freitag, Fèlix; Navarro Moldes, Leandro; Chao, Isaac; Ardáiz Villanueva, Óscar Abstract: Resource allocation is one of the challenges for self-management of large scale distributed applications running in a dynamic and heterogeneous environment. Considering Application Layer Networks (ALN) as a general term for such applications including computational Grids, Content Distribution Networks and P2P applications, the characteristics of the ALNs and the environment preclude an efficient resource allocation by a central instance. The approach we propose integrates ideas from decentralized economic models into the architecture of a resource allocation middleware, which allows the scalability towards the participant number and the robustness in very dynamic environments. At the same time, the pursuit of the participants for their individual goals should benefit the global optimization of the application. In this work, we describe the components of this middleware architecture and introduce an ongoing prototype. Wed, 05 Nov 2008 11:57:18 GMT http://hdl.handle.net/2117/2344 2008-11-05T11:57:18Z Chacín Martínez, Pablo; Freitag, Fèlix; Navarro Moldes, Leandro; Chao, Isaac; Ardáiz Villanueva, Óscar no Application layer network, Autonomic systems, Decentralized economic models, Middleware, Resource allocation, Resource self-management, Self-adjusting systems, ALN, CATNETS, P2P Resource allocation is one of the challenges for self-management of large scale distributed applications running in a dynamic and heterogeneous environment. Considering Application Layer Networks (ALN) as a general term for such applications including computational Grids, Content Distribution Networks and P2P applications, the characteristics of the ALNs and the environment preclude an efficient resource allocation by a central instance. The approach we propose integrates ideas from decentralized economic models into the architecture of a resource allocation middleware, which allows the scalability towards the participant number and the robustness in very dynamic environments. At the same time, the pursuit of the participants for their individual goals should benefit the global optimization of the application. In this work, we describe the components of this middleware architecture and introduce an ongoing prototype. http://hdl.handle.net/2117/2343 Title: Elaborating a decentralized market information system Authors: Brunner, René; Freitag, Fèlix Abstract: A Decentralized Market Information System (DMIS) that aggregates and provides information about markets is an important component for achieving markets in Grid and Peer-to-Peer systems. The proposed work is the development of a framework for the DMIS, which fulfils the economic provision within the main technical requirements like scalability towards nodes and data attributes and robustness against failures. The proposed work also allows obtaining results concerning the trade-off between economic benefits and technical costs. Introducing dynamic adaptive processes promises improvements in efficiency with regards to distributed queries and routing structures. This research proposal presents and discusses the research questions and challenges, the current knowledge and the research methodology proposed for the development of the DMIS framework. Wed, 05 Nov 2008 11:14:15 GMT http://hdl.handle.net/2117/2343 2008-11-05T11:14:15Z Brunner, René; Freitag, Fèlix no Distributed information systems, Dynamic adaptative processes, Economic benefits, Economic provision, Grid computing, Information aggregation, Marketing data processing, Monitoring systems, Publish-subscribe, Peer-to-peer computing, DMIS A Decentralized Market Information System (DMIS) that aggregates and provides information about markets is an important component for achieving markets in Grid and Peer-to-Peer systems. The proposed work is the development of a framework for the DMIS, which fulfils the economic provision within the main technical requirements like scalability towards nodes and data attributes and robustness against failures. The proposed work also allows obtaining results concerning the trade-off between economic benefits and technical costs. Introducing dynamic adaptive processes promises improvements in efficiency with regards to distributed queries and routing structures. This research proposal presents and discusses the research questions and challenges, the current knowledge and the research methodology proposed for the development of the DMIS framework. http://hdl.handle.net/2117/2339 Title: Assessing a distributed market infrastructure for economics-based service selection Authors: Brunner, René; Chao, Isaac; Chacín Martínez, Pablo; Freitag, Fèlix; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar; Joita, Liviu; Rana, Omer Abstract: Service selection is an important issue for market-oriented Grid infrastructures. However, few results have been published on the use and evaluation of market models in deployed prototypes, making it difficult to assess their capabilities. In this paper we study the integration of an extended version of Zero Intelligence Plus (ZIP) agents in a middleware for economics-based selection of Grid services. The advantages of these agents compared to alternatives is their fairly simple messaging protocol and negotiation strategy. By deploying the middleware on several machines and running experiments we observed that services are proportionally assigned to competing traders as should be in a fair market. Furthermore, varying the environmental conditions we show that the agents are able to respond to the varying environmental constraints by adapting their market prices. Mon, 03 Nov 2008 13:34:25 GMT http://hdl.handle.net/2117/2339 2008-11-03T13:34:25Z Brunner, René; Chao, Isaac; Chacín Martínez, Pablo; Freitag, Fèlix; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar; Joita, Liviu; Rana, Omer no Automatic resource allocation, Economic-based service selection, Decentralized economic models, Distributed market infraestucture, Grid computing, Marked-oriented grid infrastructures, Middleware, Multi-agent systems, Service oriented grids, Software agents, Zero intelligence plus agents, CATNETS, ZIP agents Service selection is an important issue for market-oriented Grid infrastructures. However, few results have been published on the use and evaluation of market models in deployed prototypes, making it difficult to assess their capabilities. In this paper we study the integration of an extended version of Zero Intelligence Plus (ZIP) agents in a middleware for economics-based selection of Grid services. The advantages of these agents compared to alternatives is their fairly simple messaging protocol and negotiation strategy. By deploying the middleware on several machines and running experiments we observed that services are proportionally assigned to competing traders as should be in a fair market. Furthermore, varying the environmental conditions we show that the agents are able to respond to the varying environmental constraints by adapting their market prices. http://hdl.handle.net/2117/2337 Title: Predicting MPI buffer addresses. Authors: Freitag, Fèlix; Farreras Esclusa, Montserrat; Cortés, Toni; Labarta Mancho, Jesús José Abstract: Communication latencies have been identified as one of the performance limiting factors of message passing applications in clusters of workstations/multiprocessors. On the receiver side, message-copying operations contribute to these communication latencies. Recently, prediction of MPI messages has been proposed as part of the design of a zero message-copying mechanism. Until now, prediction was only evaluated for the next message. Predicting only the next message, however, may not be enough for real implementations, since messages do not arrive in the same order as they are requested. In this paper, we explore long-term prediction of MPI messages for the design of a zero message-copying mechanism. To achieve long-term prediction we evaluate two prediction schemes, the first based on graphs, and the second based on periodicity detection. Our experiments indicate that with both prediction schemes the buffer addresses and message sizes of several future MPI messages (up to +10) can be predicted successfully. Mon, 03 Nov 2008 12:46:18 GMT http://hdl.handle.net/2117/2337 2008-11-03T12:46:18Z Freitag, Fèlix; Farreras Esclusa, Montserrat; Cortés, Toni; Labarta Mancho, Jesús José no Buffer storage, Communication latency, Graph based prediction, Message passing applications, Message-copying operations, Message size, Multiprocessor cluster, Parallel processing, Periodicity detection, Performance limiting factors, Workstation clusters, CATNETS, MPI messages Communication latencies have been identified as one of the performance limiting factors of message passing applications in clusters of workstations/multiprocessors. On the receiver side, message-copying operations contribute to these communication latencies. Recently, prediction of MPI messages has been proposed as part of the design of a zero message-copying mechanism. Until now, prediction was only evaluated for the next message. Predicting only the next message, however, may not be enough for real implementations, since messages do not arrive in the same order as they are requested. In this paper, we explore long-term prediction of MPI messages for the design of a zero message-copying mechanism. To achieve long-term prediction we evaluate two prediction schemes, the first based on graphs, and the second based on periodicity detection. Our experiments indicate that with both prediction schemes the buffer addresses and message sizes of several future MPI messages (up to +10) can be predicted successfully. http://hdl.handle.net/2117/2325 Title: Performance measuring framework for grid market middleware Authors: Freitag, Fèlix; Chacín Martínez, Pablo; Chao, Isaac; Brunner, René; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar Abstract: Current implementations of Grid infrastructures provide frameworks which aim at achieve on-demand computing. In such a scenario, contribution and use of resources will be governed by business models. The challenge is to provide multi-level performance information which enables the participation of the different actors in such a system. In this paper we describe the performance measuring framework developed for Grid Market Middleware, a middleware which supports economic-model based selection of service-oriented Grid applications. This middleware is a distributed infrastructure, which we have implemented for providing a market of services and resources to be assigned to Grid applications. The objectives of the performance measuring framework is first to assess the behaviour of the middleware and the used economic models in a deployed system, and secondly allow the provision of metrics for the components of the middleware itself. We describe the design of the performance measuring framework, its implementation and show its capability and usefulness for our objectives by experiments. Tue, 28 Oct 2008 12:54:07 GMT http://hdl.handle.net/2117/2325 2008-10-28T12:54:07Z Freitag, Fèlix; Chacín Martínez, Pablo; Chao, Isaac; Brunner, René; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar no Multi-layer performance measurement, E-service platforms, Grid middleware architecture, CATNETS Current implementations of Grid infrastructures provide frameworks which aim at achieve on-demand computing. In such a scenario, contribution and use of resources will be governed by business models. The challenge is to provide multi-level performance information which enables the participation of the different actors in such a system. In this paper we describe the performance measuring framework developed for Grid Market Middleware, a middleware which supports economic-model based selection of service-oriented Grid applications. This middleware is a distributed infrastructure, which we have implemented for providing a market of services and resources to be assigned to Grid applications. The objectives of the performance measuring framework is first to assess the behaviour of the middleware and the used economic models in a deployed system, and secondly allow the provision of metrics for the components of the middleware itself. We describe the design of the performance measuring framework, its implementation and show its capability and usefulness for our objectives by experiments. http://hdl.handle.net/2117/2324 Title: A Hayekian self-organization approach to service allocation in computing systems Authors: Eymann, Torsten; Reinicke, Michael; Freitag, Fèlix; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar; Artigas Vidal, Pau Abstract: Future ‘on-demand’ computing systems, often depicted as potentially large scale and complex Service-Oriented Architectures, will need innovative management approaches for controlling and matching services demand and supply. Centralized optimization approaches reach their bounds with increasing network size and number of nodes. The search for decentralized approaches has led to build on self-organization concepts like Autonomic Computing, which draw their inspiration from Biology. This article shows how an alternative self-organization concept from Economics, the Catallaxy concept of F.A. von Hayek, can be realized for allocating service supply and demand in a distributed "on-demand" web services network. Its implementation using a network simulator allows evaluating the approach against a centralized resource broker, by dynamically varying connection reliability and node density in the network. Exhibiting Autonomic Computing properties, the Catallaxy realization outperforms a centralized broker in highly dynamic environments. Tue, 28 Oct 2008 12:10:10 GMT http://hdl.handle.net/2117/2324 2008-10-28T12:10:10Z Eymann, Torsten; Reinicke, Michael; Freitag, Fèlix; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar; Artigas Vidal, Pau no Agent-based computational economics, Autonomic computing, Catallaxy, Digital business agents, Gride-like application layer network, Service-oriented architecture, Resource brokering, CATNET Future ‘on-demand’ computing systems, often depicted as potentially large scale and complex Service-Oriented Architectures, will need innovative management approaches for controlling and matching services demand and supply. Centralized optimization approaches reach their bounds with increasing network size and number of nodes. The search for decentralized approaches has led to build on self-organization concepts like Autonomic Computing, which draw their inspiration from Biology. This article shows how an alternative self-organization concept from Economics, the Catallaxy concept of F.A. von Hayek, can be realized for allocating service supply and demand in a distributed "on-demand" web services network. Its implementation using a network simulator allows evaluating the approach against a centralized resource broker, by dynamically varying connection reliability and node density in the network. Exhibiting Autonomic Computing properties, the Catallaxy realization outperforms a centralized broker in highly dynamic environments. http://hdl.handle.net/2117/2323 Title: A catallactic market for data mining services. Authors: Joita, Liviu; Rana, Omer; Freitag, Fèlix; Chao, Isaac; Chacín Martínez, Pablo; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar Abstract: We describe a Grid market for exchanging data mining services based on the Catallactic market mechanism proposed by von Hayek. This market mechanism allows selection between multiple instances of services based on operations required in a data mining task (such as data migration, data pre-processing and subsequently data analysis). Catallaxy is a decentralized approach, based on a “free market” mechanism, and is particularly useful when the number of market participants is large or when conditions within the market often change. It is therefore particularly suitable in Grid and peer-2-peer systems. The approach assumes that the service provider and user are not co-located, and require multiple message exchanges to carry out a data mining task. A market of J48-based decision tree algorithm instances, each implemented as a Web service, is used to demonstrate our approach. We have validated the feasibility of building catallactic data mining grid applications, and implemented a proof-of-concept application (Cat-COVITE) mapped to a Catallactic Grid Middleware. Mon, 27 Oct 2008 13:01:18 GMT http://hdl.handle.net/2117/2323 2008-10-27T13:01:18Z Joita, Liviu; Rana, Omer; Freitag, Fèlix; Chao, Isaac; Chacín Martínez, Pablo; Navarro Moldes, Leandro; Ardáiz Villanueva, Óscar no Data mining, Grid market, Midleware, P-2-P systems, CatCOVITE, CATNETS We describe a Grid market for exchanging data mining services based on the Catallactic market mechanism proposed by von Hayek. This market mechanism allows selection between multiple instances of services based on operations required in a data mining task (such as data migration, data pre-processing and subsequently data analysis). Catallaxy is a decentralized approach, based on a “free market” mechanism, and is particularly useful when the number of market participants is large or when conditions within the market often change. It is therefore particularly suitable in Grid and peer-2-peer systems. The approach assumes that the service provider and user are not co-located, and require multiple message exchanges to carry out a data mining task. A market of J48-based decision tree algorithm instances, each implemented as a Web service, is used to demonstrate our approach. We have validated the feasibility of building catallactic data mining grid applications, and implemented a proof-of-concept application (Cat-COVITE) mapped to a Catallactic Grid Middleware. http://hdl.handle.net/2117/2306 Title: Group-awareness for mobile cooperative learning. Authors: Meseguer Pallarès, Roc; Navarro Moldes, Leandro; Reyes Muñoz, María Angélica Abstract: Collaborative learning traditionally relies on the arrangement of space and physical artifacts to mediate group activities. Our claim is that CSCL environments should enhance collaborative learning in a complementary way without introducing additional burden to human participants. This paper present research results on how tracking the location of people could be used to incorporate automatic group awareness in existing and new CSCL applications that support mobile cooperative learning. After presenting the requirements and structure of a middleware to handle automatic group awareness based on location information, two applications are presented. Experiments in a real setting on a jigsaw activity, with students using laptops with WiFi on a university course, confirm the validity of using location technology to automatically derive group structure and the positive effect on the learning results. Wed, 15 Oct 2008 10:28:57 GMT http://hdl.handle.net/2117/2306 2008-10-15T10:28:57Z Meseguer Pallarès, Roc; Navarro Moldes, Leandro; Reyes Muñoz, María Angélica no Automatic group awareness, Collaborative learning, Computer-aided instruction, Jigsaw activity, Middleware, Mobile computing, Mobile cooperative learning, CSCL Collaborative learning traditionally relies on the arrangement of space and physical artifacts to mediate group activities. Our claim is that CSCL environments should enhance collaborative learning in a complementary way without introducing additional burden to human participants. This paper present research results on how tracking the location of people could be used to incorporate automatic group awareness in existing and new CSCL applications that support mobile cooperative learning. After presenting the requirements and structure of a middleware to handle automatic group awareness based on location information, two applications are presented. Experiments in a real setting on a jigsaw activity, with students using laptops with WiFi on a university course, confirm the validity of using location technology to automatically derive group structure and the positive effect on the learning results. http://hdl.handle.net/2117/2304 Title: WWG: a wide-area infrastructure for group work Authors: Marquès Puig, Joan Manel; Navarro Moldes, Leandro Abstract: Group learning at Internet scale is becoming more frequent in university courses. This complex process requires support by distributed computing learning infrastructures. This paper describes the design of WWG: a distributed and decentralized infrastructure with the aim of supporting distributed group learning and team work, centered on the distribution of events, so that every participant can be notified and thus be aware of the actions, changes, progress of the groups he or she belongs to - synchronous awareness for asynchronous work. The design issues, requirements and the resulting architecture are presented. WWG is based on a multicast mechanism for event distribution with meta-information agents responsible for the dissemination and transformation of events, repository agents responsible for the storage of group information and user agents responsible for the representation of users (sources and sinks of events) Tue, 14 Oct 2008 13:02:17 GMT http://hdl.handle.net/2117/2304 2008-10-14T13:02:17Z Marquès Puig, Joan Manel; Navarro Moldes, Leandro no Decentralized infrastructure, Distance learning, Distributed group learning, Distributed object management, Event distribution, Groupware, Internet-scale distributed systems, Middleware, University courses, Virtual groups, Wide-area infrastructure, WWG Group learning at Internet scale is becoming more frequent in university courses. This complex process requires support by distributed computing learning infrastructures. This paper describes the design of WWG: a distributed and decentralized infrastructure with the aim of supporting distributed group learning and team work, centered on the distribution of events, so that every participant can be notified and thus be aware of the actions, changes, progress of the groups he or she belongs to - synchronous awareness for asynchronous work. The design issues, requirements and the resulting architecture are presented. WWG is based on a multicast mechanism for event distribution with meta-information agents responsible for the dissemination and transformation of events, repository agents responsible for the storage of group information and user agents responsible for the representation of users (sources and sinks of events) http://hdl.handle.net/2117/2303 Title: P2P architecture for scientific collaboration. Authors: Mitre Silva, José; Navarro Moldes, Leandro Abstract: P2P networks are often associated with file exchange applications among private users. However, their features make them suitable for other uses. In this paper we present a P2P architecture for scientific collaboration networks, which takes advantage of the properties inherent in these social networks - small-world, clustering, community structure, assortative mixing, preferential attachment and small and stable groups - in order to obtain better performance, efficient use of resources and system resilience. Tue, 14 Oct 2008 11:17:05 GMT http://hdl.handle.net/2117/2303 2008-10-14T11:17:05Z Mitre Silva, José; Navarro Moldes, Leandro no Associative mixing, Degree distribution, File exchange, Groupware, Peer-to-peer computing, Preferential attachmet, Resource allocation, Scientific collaboration networks, Social networks, System resilience, P2P architecture P2P networks are often associated with file exchange applications among private users. However, their features make them suitable for other uses. In this paper we present a P2P architecture for scientific collaboration networks, which takes advantage of the properties inherent in these social networks - small-world, clustering, community structure, assortative mixing, preferential attachment and small and stable groups - in order to obtain better performance, efficient use of resources and system resilience. http://hdl.handle.net/2117/2302 Title: Extending the scope of asynchronous collaboration: a matter of being autonomous and self-sufficient Authors: Marquès Puig, Joan Manel; Navarro Moldes, Leandro; Daradoumis, Athanasios Abstract: Asynchronous collaborative applications and systems have to deal with complexities associated with interaction nature, idiosyncrasy of groups and technical and administrative issues. Inclusion of requirements derived from them is costly (in time, resources and economically). Existing solutions addresses asynchronous collaboration via simplification of requirements and by using centralized models. In this paper we present LaCOLLA, a fully decentralized infrastructure for building collaborative applications that provides general purpose collaborative functionalities. The provision of those functionalities will avoid applications deal with most of complexities derived from groups and its members, what will help inclusion of collaborative aspects. The implementation of LaCOLLA follows the peer-to-peer paradigm and pays special attention to autonomy of its members and to self-organization of the components of the infrastructure. Another key aspect is that resources (e.g. storage) and services (e.g. authorization) are provided by its members (avoiding dependency from agents not belonging to group). Tue, 14 Oct 2008 11:00:06 GMT http://hdl.handle.net/2117/2302 2008-10-14T11:00:06Z Marquès Puig, Joan Manel; Navarro Moldes, Leandro; Daradoumis, Athanasios no Asynchronous collaboration, Decentralized collaborative infrastructures, Groupware, Internet, Middleware, Peer-to-peer computing, Resource allocation, LaCOLLA Asynchronous collaborative applications and systems have to deal with complexities associated with interaction nature, idiosyncrasy of groups and technical and administrative issues. Inclusion of requirements derived from them is costly (in time, resources and economically). Existing solutions addresses asynchronous collaboration via simplification of requirements and by using centralized models. In this paper we present LaCOLLA, a fully decentralized infrastructure for building collaborative applications that provides general purpose collaborative functionalities. The provision of those functionalities will avoid applications deal with most of complexities derived from groups and its members, what will help inclusion of collaborative aspects. The implementation of LaCOLLA follows the peer-to-peer paradigm and pays special attention to autonomy of its members and to self-organization of the components of the infrastructure. Another key aspect is that resources (e.g. storage) and services (e.g. authorization) are provided by its members (avoiding dependency from agents not belonging to group). http://hdl.handle.net/2117/2301 Title: Xweb: a framework for application network deployment in a programmable internet service infrastructure Authors: Ardáiz Villanueva, Óscar; Navarro Moldes, Leandro Abstract: An application network consists of a number of application servers distributed throughout the Internet, connected and coordinated to provide services with low latency. Adding, removing and migrating servers, application networks adapt to demand variations. To create new servers anywhere in the Internet a programmable Internet service infrastructure is needed. In addition application network servers must be deployed coordinately. We propose a framework for application network deployment that implements such functionality. Tue, 14 Oct 2008 10:30:12 GMT http://hdl.handle.net/2117/2301 2008-10-14T10:30:12Z Ardáiz Villanueva, Óscar; Navarro Moldes, Leandro no Application network deployment, Application network servers, Client-server systems, Programable internet, Service infrastructures, Xweb framework An application network consists of a number of application servers distributed throughout the Internet, connected and coordinated to provide services with low latency. Adding, removing and migrating servers, application networks adapt to demand variations. To create new servers anywhere in the Internet a programmable Internet service infrastructure is needed. In addition application network servers must be deployed coordinately. We propose a framework for application network deployment that implements such functionality. http://hdl.handle.net/2117/2299 Title: Multicast injection for application network deployment. Authors: Ardáiz Villanueva, Óscar; Freitag, Fèlix; Navarro Moldes, Leandro Abstract: Introduction of new services on the Internet is a laborious, time-consuming task. Application networks, applications being serviced through multiple interconnected service nodes disseminated across the Internet for better performance, fault tolerance and availability, as well are costly to set up. In order to provide a network-enabled application service, a number of surrogate servers have to be provisioned. in this paper we propose a mechanisms to dynamically deploy an application network: multicast injection. Currently employed dynamic deployment mechanisms, SNMP per-node configuration, is a centralized model that can not scale or be as fault tolerant as more distributed mechanisms such as multicast injection. We perform simulations to compare its efficiency in terms of deployment request success ratio, unused allocation percentage and traffic vs. deployment resource allocation requests. We show that multicast injection has a higher success ratio with lower bandwidth consumption at the deployers' locations. Mon, 13 Oct 2008 15:44:09 GMT http://hdl.handle.net/2117/2299 2008-10-13T15:44:09Z Ardáiz Villanueva, Óscar; Freitag, Fèlix; Navarro Moldes, Leandro no Bandwidth consumption, Computer network reliability, Dynamic development mechanism, Faul tolerance, Multicast communications, Network servers, Performance evaluation, SNMP per-mode configuration, Telecommunications traffic, Unused allocation percentage Introduction of new services on the Internet is a laborious, time-consuming task. Application networks, applications being serviced through multiple interconnected service nodes disseminated across the Internet for better performance, fault tolerance and availability, as well are costly to set up. In order to provide a network-enabled application service, a number of surrogate servers have to be provisioned. in this paper we propose a mechanisms to dynamically deploy an application network: multicast injection. Currently employed dynamic deployment mechanisms, SNMP per-node configuration, is a centralized model that can not scale or be as fault tolerant as more distributed mechanisms such as multicast injection. We perform simulations to compare its efficiency in terms of deployment request success ratio, unused allocation percentage and traffic vs. deployment resource allocation requests. We show that multicast injection has a higher success ratio with lower bandwidth consumption at the deployers' locations. http://hdl.handle.net/2117/2297 Title: Jacobi orderings for multi-port hypercubes. Authors: Royo Vallés, María Dolores; González Colás, Antonio María; Valero García, Miguel Abstract: The communication cost plays a key role in the performance of many parallel algorithms. In the particular case of the one-sided Jacobi method for symmetric eigenvalue and eigenvector computation the communication cost of previously proposed algorithms is mainly determined by the particular ordering being used. We propose two novel Jacobi orderings: the permuted-BR ordering and the degree-4 ordering, aimed at efficiently exploiting the multi-port capability of a hypercube. It is shown that the former is nearly optimal for some scenarios and the latter outperforms previously known orderings by a factor of two. Mon, 13 Oct 2008 15:16:39 GMT http://hdl.handle.net/2117/2297 2008-10-13T15:16:39Z Royo Vallés, María Dolores; González Colás, Antonio María; Valero García, Miguel no Communication cost, Degree-4 ordering, Eigenvalue computation, Hypercube networks, Jacobian matrices, Jacobian orderings, Mathematics computing, Multiport hypercubes, One-sided Jacobi method, Parallel algoriths, Parallel architectures, Permuted-BR ordering The communication cost plays a key role in the performance of many parallel algorithms. In the particular case of the one-sided Jacobi method for symmetric eigenvalue and eigenvector computation the communication cost of previously proposed algorithms is mainly determined by the particular ordering being used. We propose two novel Jacobi orderings: the permuted-BR ordering and the degree-4 ordering, aimed at efficiently exploiting the multi-port capability of a hypercube. It is shown that the former is nearly optimal for some scenarios and the latter outperforms previously known orderings by a factor of two.