This paper describes the design, implementation, and evaluation of a parallel indexer called PAMI... more This paper describes the design, implementation, and evaluation of a parallel indexer called PAMIS ¢ f or a polygonal 2D shape image database. PAMIS is based on a shape representation scheme called the turning function, which exhibits the desirable properties of position-, scale-, and rotation-invariance, and has a similarity metric function that satisfies the triangular inequality, which is required for efficient database indexing. Because the goal of the PAMIS project is to support "like-this" image queries, the indexing scheme we chose, the vantage-point tree (VPT), uses relative rather than absolute distance values to organize the database elements for efficient nearest-neighbor search. We have successfully implemented PAMIS on a network of workstations to exploit the I/O and computation parallelism inherent in the VPT algorithm. We found that it is preferable to make the VPT node size as small as possible in order to have a lean and deep VPT structure, and the best-case scheduling strategy performs the best among the scheduling strategies considered. Overall, the performance of the VPT algorithm scales very well with the number of processors, and the indexing efficiency, defined as the percentage of database elements touched by the search, of PAMIS is 6% and 39% for "good" queries that ask for 1 and 50 nearest neighbors, respectively.
Scientific codes are increasingly being used in compositional settings, especially problem solvin... more Scientific codes are increasingly being used in compositional settings, especially problem solving environments (PSEs). Typical compositional modeling frameworks require significant buy-in, in the form of commitment to a particular style of programming (e.g., distributed object components). While this solution is feasible for newer generations of component-based scientific codes, large legacy code bases present a veritable software engineering nightmare. We introduce Weaves-a novel framework that enables modeling, composition, direct code execution, performance characterization, adaptation, and control of unmodified high performance scientific codes. Weaves is an efficient generalized framework for parallel compositional modeling that is a proper superset of the threads and processes models of programming. In this paper, our focus is on the transparent code execution interface enabled by Weaves. We identify design constraints, their impact on implementation alternatives, configuration scenarios, and present results from a prototype implementation on Intel x86 architectures.
In this paper we describe a distributed-memory parallel implementation of the Open Network Emulat... more In this paper we describe a distributed-memory parallel implementation of the Open Network Emulator (ONE), a network simulator that combines the controllability of simulation with the direct code execution advantages of emulation and experimental testbeds. ONE uses a scaled real-time model called Relativistic Time (RT). We describe a RT-based reactive global warp detection algorithm to exploit lookahead. The Distributed ONE system provides a platform for network simulation that combines model fidelity (existing network applications can be compiled and instantiated within ONE without modification), temporal fidelity, and good scalability. We present strong and weak scaling performance results for the Distributed ONE system on up to sixteen nodes of a distributed-memory parallel cluster, on simulations involving up to 16,000 virtual hosts.
L'invention concerne un systeme reparti utilise pour creer un point de controle pour une plur... more L'invention concerne un systeme reparti utilise pour creer un point de controle pour une pluralite de processus executes dans le systeme reparti. Le systeme reparti comprend plusieurs noeuds de calcul, un systeme d'exploitation s'executant dans chaque noeud de calcul. Une bibliotheque de points de controle residant au niveau de l'utilisateur, dans chaque noeud de calcul, est transparente pour le systeme d'exploitation residant dans le meme noeud de calcul et pour les autres noeuds de calcul. Chaque bibliotheque de points de controle utilise un protocole d'entree en session de messagerie par fenetrage pour la creation des points de controle du systeme reparti. Des processus participant a un calcul distribue dans le systeme reparti peuvent etre migres d'un noeud de calcul a un autre noeud de calcul du systeme reparti, par remappage d'adresses de materiel a l'aide de la bibliotheque de points de controle.
This dissertation proposes the Weaves runtime framework for the execution of large scale parallel... more This dissertation proposes the Weaves runtime framework for the execution of large scale parallel programs over lightweight intra-process threads. The goal of the Weaves framework is to help process-based legacy parallel programs exploit the scalability of threads without any modifications. The framework separates global variables used by identical, but independent, threads of legacy parallel programs without resorting to thread-based reprogramming. At the same time, it also facilitates low-overhead collaboration among threads of a legacy parallel program through multi-granular selective sharing of global variables. Applications that follow the tenets of the Weaves framework can load multiple identical, but independent, copies of arbitrary object files within a single process. They can compose the runtime images of these object files in graph-like ways and run intra-process threads through them to realize various degrees of multi-granular selective sharing or separation of global variables among the threads. Using direct runtime control over the resolution of individual references to functions and variables, they can also manipulate program composition at fine granularities. Most importantly, the Weaves framework does not entail any modifications to either the source codes or the native codes of the object files. The framework is completely transparent. Results from experiments with a real-world process-based parallel application depict that I take this opportunity to thank the people who helped me in various ways throughout the time-span of this work: My advisor, Dr. Srinidhi Varadarajan, who has been a friend, a philosopher and a guide. He helped me at all stages of this work, from conceptualization to implementation.
Uncovering the metrics and procedures employed by an autonomous networking system is an important... more Uncovering the metrics and procedures employed by an autonomous networking system is an important problem with applications in instrumentation, traffic engineering, and game-theoretic studies of multi-agent environments. This thesis presents a method for utilizing inverse reinforcement learning (IRL) techniques for the purpose of discovering a composite metric used by a dynamic routing algorithm on an Internet Protocol (IP) network. The network and routing algorithm are modeled as a reinforcement learning (RL) agent and a Markov decision process (MDP). The problem of routing metric discovery is then posed as a problem of recovering the reward function, given observed optimal behavior. We show that this approach is empirically suited for determining the relative contributions of factors that constitute a composite metric. Experimental results for many classes of randomly generated networks are presented. Acknowledgments I would like to thank my advisers Dr. Naren Ramakrishnan and Dr....
This paper studies the evaluation of routing algorithms from the perspective of reachability rout... more This paper studies the evaluation of routing algorithms from the perspective of reachability routing, where the goal is to determine all paths between a sender and a receiver. Reachability routing is becoming relevant with the changing dynamics of the Internet and the emergence of low-bandwidth wireless/ad-hoc networks. We make the case for reinforcement learning as the framework of choice to realize reachability routing, within the confines of the current Internet infrastructure. The setting of the reinforcement learning problem offers several advantages, including loop resolution, multi-path forwarding capability, cost-sensitive routing, and minimizing state overhead, while maintaining the incremental spirit of current backbone routing algorithms. We identify research issues in reinforcement learning applied to the reachability routing problem to achieve a fluid and robust backbone routing framework. The paper is targeted toward practitioners seeking to implement a reachability ro...
As supercomputers continue to grow in scale and capabilities, it is becoming increasingly difficu... more As supercomputers continue to grow in scale and capabilities, it is becoming increasingly difficult to isolate processor and system level causes of performance degradation. Over the last several years, a significant number of performance analysis and monitoring tools have been built/proposed. However, these tools suffer from several important shortcomings, particularly in distributed environments. In this paper we present ScALPEL, a Scalable Adaptive Lightweight Performance Evaluation Library for application performance monitoring at the functional level. Our approach provides several distinct advantages. First, ScALPEL is portable across a wide variety of architectures, and its ability to selectively monitor functions presents low run-time overhead, enabling its use for large-scale production applications. Second, it is run-time configurable, enabling both dynamic selection of functions to profile as well as events of interest on a per function basis. Third, our approach is transpa...
We present a modular approach to realizing fine-grained adaptation of program behavior in a paral... more We present a modular approach to realizing fine-grained adaptation of program behavior in a parallel environment. Using a compositional framework based on function call interception and manipulation, the adaptive logic to monitor internal program states and control the behavior of program modules is written and managed as a separate code, thus supporting centralized design of complex adaptation strategies for adapting to dynamic changes within an application. By 'catching' the functions that execute in synchronization across the parallel environment and inserting the adaptive logic operations at the intercepted control points, the proposed method provides a convenient way of synchronous adaptation without disturbing the parallel execution and communication structure already established in the original program. Applying our method to a CFD (computational fluid dynamics) simulation program to implement example adaptation scenarios, we demonstrate how effectively applications can change their behavior through fine-grained control.
Scientific software often needs to be adapted for different execution environments, problem sets,... more Scientific software often needs to be adapted for different execution environments, problem sets, and available resources to ensure its efficiency and reliability. However, for existing programs, implementing adaptations by directly modifying source code can be time-consuming, error-prone, and difficult to manage for today's complex software. In this paper, we present a modular approach to realizing adaptation for existing scientific codes. By treating adaptation as a separate concern, our approach supports the development of application-specific, context-aware adaptation schemes without directly modifying the original code. Our approach uses a compositional framework that offers language-neutral mechanisms to integrate separately written adaptation code with existing code. Using our approach, scientific programmers can focus on the design and implementation of adaptation schemes separately from the original code development, and then compose an adaptive application whose original capabilities are enhanced in diverse aspects such as performance and stability. Our compositional approach enables fine-grained adaptation, so that an application's program behavior is controlled at the function or algorithm level by adaptation code plugged into the application. By applying our approach to real-world scientific applications to implement various adaptation scenarios, we demonstrate applicability and effectiveness for adapting scientific software.
Keyword-based text retrieval engines have been and will continue to be essential to text-based in... more Keyword-based text retrieval engines have been and will continue to be essential to text-based information access systems because they serve as the basic building blocks to high-level text analysis systems. Traditionally, text compression and text retrieval are teated as independent problems. Text les are compressed and indexed separately. To answer a keyword-based query, text les are rst uncompressed, and then searched sequentially or via an inverted index. This paper describes the design, implementation and evaluation of a novel integrated text compression and indexing scheme called ITCI, which combines the dictionary data structures for compression and indexing, and allows direct search through compressed text. The performance results show that ITCI's compression eeciency is within 7% to 17% of GZIP, which is among the best lossless data compression algorithms, The sum of the compressed text and the inverted index is only between 55% to 76% of the original text size, while th...
Keyword based search engines are the basic building block of text retrieval systems. Higher level... more Keyword based search engines are the basic building block of text retrieval systems. Higher level systems like content sensitive search engines and knowledge-based systems still rely on keyword search as the underlying text retrieval mechanism. With the explosive growth in content, Internet and Intranet information repositories require efficient mechanisms to store as well as index data. In this paper we discuss the implementation of the Shrink and Search Engine (SASE) framework which unites text compression and indexing to maximize keyword search performance while reducing storage cost. SASE features the novel capability of being able to directly search through compressed text without explicit decompression. The implementation includes a search server architecture, which can be accessed from a Java front-end to perform keyword search on the Internet. The performance results show that the compression efficiency of SASE is within 7-17% of GZIP one of the best lossless compression sch...
We recommend a programming construct - availability check - for programs that need to automatical... more We recommend a programming construct - availability check - for programs that need to automatically adjust to presence or absence of segments of code. The idea is to check the existence of a valid definition before a function call is invoked. The syntax is that of a simple 'if' statement. The vision is to enable customization of application functionality through addition or removal of optional components, but without requiring complete re-building. Focus is on C-like compiled procedural languages and UNIX-based systems. Essentially, our approach attempts to combine the flexibility of dynamic libraries with the usability of utility (dependency) libraries. We outline the benefits over prevalent strategies mainly in terms of development complexity, crudely measured as lesser lines of code. We also allude to performance and flexibility facets. A Preliminary implementation and figures from early experimental evaluation are presented.
Current directions in network routing research have not kept pace with the latest developments in... more Current directions in network routing research have not kept pace with the latest developments in network architectures, such as peer-to-peer networks, sensor networks, ad-hoc wireless networks, and overlay networks. A common characteristic among all of these new technologies is the presence of highly dynamic network topologies. Currently deployed single-path routing protocols cannot adequately cope with this dynamism, and existing multi-path algorithms make trade-offs which lead to less than optimal performance on these networks. This drives the need for routing protocols designed with the unique characteristics of these networks in mind. In this paper we propose the notion of reachability routing as a solution to the challenges posed by routing on such dynamic networks. In particular, our formulation of reachability routing provides cost-sensitive multi-path forwarding along with loop avoidance within the confines of the Internet Protocol (IP) architecture. This is achieved throug...
In this paper, we illustrate the impact of dynamic resizability on parallel scheduling. Our ReSHA... more In this paper, we illustrate the impact of dynamic resizability on parallel scheduling. Our ReSHAPE framework includes an application scheduler that supports dynamic resizing of parallel applications. We propose and evaluate new scheduling policies made possible by our ReSHAPE framework. The framework also provides a platform to experiment with more interesting and sophisticated scheduling policies and scenarios for resizable parallel applications. The proposed policies support scheduling of parallel applications with and without user assigned priorities. Experimental results show that these scheduling policies significantly improve individual application turn around time as well as overall cluster utilization.
Grid infrastructures and computing environments have progressed significantly in the past few yea... more Grid infrastructures and computing environments have progressed significantly in the past few years. The vision of truly seamless Grid usage relies on runtime systems support that is cognizant of the operational issues underlying grid computations and, at the same time, is flexible enough to accommodate diverse application scenarios. This paper addresses the twin aspects of Grid infrastructure and application support through a novel combination of two computational technologies – Weaves, a source-language independent parallel runtime compositional framework that operates through reverse-analysis of compiled object files, and runtime recommender systems that aid in dynamic knowledge-based application composition. Domain-specific adaptivity is exploited through a novel compositional system that supports runtime recommendation of code modules and a sophisticated checkpointing and runtime migration solution that can be transparently deployed over Grid infrastructures. A core set of “ada...
This paper describes the design, implementation, and evaluation of a parallel indexer called PAMI... more This paper describes the design, implementation, and evaluation of a parallel indexer called PAMIS ¢ f or a polygonal 2D shape image database. PAMIS is based on a shape representation scheme called the turning function, which exhibits the desirable properties of position-, scale-, and rotation-invariance, and has a similarity metric function that satisfies the triangular inequality, which is required for efficient database indexing. Because the goal of the PAMIS project is to support "like-this" image queries, the indexing scheme we chose, the vantage-point tree (VPT), uses relative rather than absolute distance values to organize the database elements for efficient nearest-neighbor search. We have successfully implemented PAMIS on a network of workstations to exploit the I/O and computation parallelism inherent in the VPT algorithm. We found that it is preferable to make the VPT node size as small as possible in order to have a lean and deep VPT structure, and the best-case scheduling strategy performs the best among the scheduling strategies considered. Overall, the performance of the VPT algorithm scales very well with the number of processors, and the indexing efficiency, defined as the percentage of database elements touched by the search, of PAMIS is 6% and 39% for "good" queries that ask for 1 and 50 nearest neighbors, respectively.
Scientific codes are increasingly being used in compositional settings, especially problem solvin... more Scientific codes are increasingly being used in compositional settings, especially problem solving environments (PSEs). Typical compositional modeling frameworks require significant buy-in, in the form of commitment to a particular style of programming (e.g., distributed object components). While this solution is feasible for newer generations of component-based scientific codes, large legacy code bases present a veritable software engineering nightmare. We introduce Weaves-a novel framework that enables modeling, composition, direct code execution, performance characterization, adaptation, and control of unmodified high performance scientific codes. Weaves is an efficient generalized framework for parallel compositional modeling that is a proper superset of the threads and processes models of programming. In this paper, our focus is on the transparent code execution interface enabled by Weaves. We identify design constraints, their impact on implementation alternatives, configuration scenarios, and present results from a prototype implementation on Intel x86 architectures.
In this paper we describe a distributed-memory parallel implementation of the Open Network Emulat... more In this paper we describe a distributed-memory parallel implementation of the Open Network Emulator (ONE), a network simulator that combines the controllability of simulation with the direct code execution advantages of emulation and experimental testbeds. ONE uses a scaled real-time model called Relativistic Time (RT). We describe a RT-based reactive global warp detection algorithm to exploit lookahead. The Distributed ONE system provides a platform for network simulation that combines model fidelity (existing network applications can be compiled and instantiated within ONE without modification), temporal fidelity, and good scalability. We present strong and weak scaling performance results for the Distributed ONE system on up to sixteen nodes of a distributed-memory parallel cluster, on simulations involving up to 16,000 virtual hosts.
L'invention concerne un systeme reparti utilise pour creer un point de controle pour une plur... more L'invention concerne un systeme reparti utilise pour creer un point de controle pour une pluralite de processus executes dans le systeme reparti. Le systeme reparti comprend plusieurs noeuds de calcul, un systeme d'exploitation s'executant dans chaque noeud de calcul. Une bibliotheque de points de controle residant au niveau de l'utilisateur, dans chaque noeud de calcul, est transparente pour le systeme d'exploitation residant dans le meme noeud de calcul et pour les autres noeuds de calcul. Chaque bibliotheque de points de controle utilise un protocole d'entree en session de messagerie par fenetrage pour la creation des points de controle du systeme reparti. Des processus participant a un calcul distribue dans le systeme reparti peuvent etre migres d'un noeud de calcul a un autre noeud de calcul du systeme reparti, par remappage d'adresses de materiel a l'aide de la bibliotheque de points de controle.
This dissertation proposes the Weaves runtime framework for the execution of large scale parallel... more This dissertation proposes the Weaves runtime framework for the execution of large scale parallel programs over lightweight intra-process threads. The goal of the Weaves framework is to help process-based legacy parallel programs exploit the scalability of threads without any modifications. The framework separates global variables used by identical, but independent, threads of legacy parallel programs without resorting to thread-based reprogramming. At the same time, it also facilitates low-overhead collaboration among threads of a legacy parallel program through multi-granular selective sharing of global variables. Applications that follow the tenets of the Weaves framework can load multiple identical, but independent, copies of arbitrary object files within a single process. They can compose the runtime images of these object files in graph-like ways and run intra-process threads through them to realize various degrees of multi-granular selective sharing or separation of global variables among the threads. Using direct runtime control over the resolution of individual references to functions and variables, they can also manipulate program composition at fine granularities. Most importantly, the Weaves framework does not entail any modifications to either the source codes or the native codes of the object files. The framework is completely transparent. Results from experiments with a real-world process-based parallel application depict that I take this opportunity to thank the people who helped me in various ways throughout the time-span of this work: My advisor, Dr. Srinidhi Varadarajan, who has been a friend, a philosopher and a guide. He helped me at all stages of this work, from conceptualization to implementation.
Uncovering the metrics and procedures employed by an autonomous networking system is an important... more Uncovering the metrics and procedures employed by an autonomous networking system is an important problem with applications in instrumentation, traffic engineering, and game-theoretic studies of multi-agent environments. This thesis presents a method for utilizing inverse reinforcement learning (IRL) techniques for the purpose of discovering a composite metric used by a dynamic routing algorithm on an Internet Protocol (IP) network. The network and routing algorithm are modeled as a reinforcement learning (RL) agent and a Markov decision process (MDP). The problem of routing metric discovery is then posed as a problem of recovering the reward function, given observed optimal behavior. We show that this approach is empirically suited for determining the relative contributions of factors that constitute a composite metric. Experimental results for many classes of randomly generated networks are presented. Acknowledgments I would like to thank my advisers Dr. Naren Ramakrishnan and Dr....
This paper studies the evaluation of routing algorithms from the perspective of reachability rout... more This paper studies the evaluation of routing algorithms from the perspective of reachability routing, where the goal is to determine all paths between a sender and a receiver. Reachability routing is becoming relevant with the changing dynamics of the Internet and the emergence of low-bandwidth wireless/ad-hoc networks. We make the case for reinforcement learning as the framework of choice to realize reachability routing, within the confines of the current Internet infrastructure. The setting of the reinforcement learning problem offers several advantages, including loop resolution, multi-path forwarding capability, cost-sensitive routing, and minimizing state overhead, while maintaining the incremental spirit of current backbone routing algorithms. We identify research issues in reinforcement learning applied to the reachability routing problem to achieve a fluid and robust backbone routing framework. The paper is targeted toward practitioners seeking to implement a reachability ro...
As supercomputers continue to grow in scale and capabilities, it is becoming increasingly difficu... more As supercomputers continue to grow in scale and capabilities, it is becoming increasingly difficult to isolate processor and system level causes of performance degradation. Over the last several years, a significant number of performance analysis and monitoring tools have been built/proposed. However, these tools suffer from several important shortcomings, particularly in distributed environments. In this paper we present ScALPEL, a Scalable Adaptive Lightweight Performance Evaluation Library for application performance monitoring at the functional level. Our approach provides several distinct advantages. First, ScALPEL is portable across a wide variety of architectures, and its ability to selectively monitor functions presents low run-time overhead, enabling its use for large-scale production applications. Second, it is run-time configurable, enabling both dynamic selection of functions to profile as well as events of interest on a per function basis. Third, our approach is transpa...
We present a modular approach to realizing fine-grained adaptation of program behavior in a paral... more We present a modular approach to realizing fine-grained adaptation of program behavior in a parallel environment. Using a compositional framework based on function call interception and manipulation, the adaptive logic to monitor internal program states and control the behavior of program modules is written and managed as a separate code, thus supporting centralized design of complex adaptation strategies for adapting to dynamic changes within an application. By 'catching' the functions that execute in synchronization across the parallel environment and inserting the adaptive logic operations at the intercepted control points, the proposed method provides a convenient way of synchronous adaptation without disturbing the parallel execution and communication structure already established in the original program. Applying our method to a CFD (computational fluid dynamics) simulation program to implement example adaptation scenarios, we demonstrate how effectively applications can change their behavior through fine-grained control.
Scientific software often needs to be adapted for different execution environments, problem sets,... more Scientific software often needs to be adapted for different execution environments, problem sets, and available resources to ensure its efficiency and reliability. However, for existing programs, implementing adaptations by directly modifying source code can be time-consuming, error-prone, and difficult to manage for today's complex software. In this paper, we present a modular approach to realizing adaptation for existing scientific codes. By treating adaptation as a separate concern, our approach supports the development of application-specific, context-aware adaptation schemes without directly modifying the original code. Our approach uses a compositional framework that offers language-neutral mechanisms to integrate separately written adaptation code with existing code. Using our approach, scientific programmers can focus on the design and implementation of adaptation schemes separately from the original code development, and then compose an adaptive application whose original capabilities are enhanced in diverse aspects such as performance and stability. Our compositional approach enables fine-grained adaptation, so that an application's program behavior is controlled at the function or algorithm level by adaptation code plugged into the application. By applying our approach to real-world scientific applications to implement various adaptation scenarios, we demonstrate applicability and effectiveness for adapting scientific software.
Keyword-based text retrieval engines have been and will continue to be essential to text-based in... more Keyword-based text retrieval engines have been and will continue to be essential to text-based information access systems because they serve as the basic building blocks to high-level text analysis systems. Traditionally, text compression and text retrieval are teated as independent problems. Text les are compressed and indexed separately. To answer a keyword-based query, text les are rst uncompressed, and then searched sequentially or via an inverted index. This paper describes the design, implementation and evaluation of a novel integrated text compression and indexing scheme called ITCI, which combines the dictionary data structures for compression and indexing, and allows direct search through compressed text. The performance results show that ITCI's compression eeciency is within 7% to 17% of GZIP, which is among the best lossless data compression algorithms, The sum of the compressed text and the inverted index is only between 55% to 76% of the original text size, while th...
Keyword based search engines are the basic building block of text retrieval systems. Higher level... more Keyword based search engines are the basic building block of text retrieval systems. Higher level systems like content sensitive search engines and knowledge-based systems still rely on keyword search as the underlying text retrieval mechanism. With the explosive growth in content, Internet and Intranet information repositories require efficient mechanisms to store as well as index data. In this paper we discuss the implementation of the Shrink and Search Engine (SASE) framework which unites text compression and indexing to maximize keyword search performance while reducing storage cost. SASE features the novel capability of being able to directly search through compressed text without explicit decompression. The implementation includes a search server architecture, which can be accessed from a Java front-end to perform keyword search on the Internet. The performance results show that the compression efficiency of SASE is within 7-17% of GZIP one of the best lossless compression sch...
We recommend a programming construct - availability check - for programs that need to automatical... more We recommend a programming construct - availability check - for programs that need to automatically adjust to presence or absence of segments of code. The idea is to check the existence of a valid definition before a function call is invoked. The syntax is that of a simple 'if' statement. The vision is to enable customization of application functionality through addition or removal of optional components, but without requiring complete re-building. Focus is on C-like compiled procedural languages and UNIX-based systems. Essentially, our approach attempts to combine the flexibility of dynamic libraries with the usability of utility (dependency) libraries. We outline the benefits over prevalent strategies mainly in terms of development complexity, crudely measured as lesser lines of code. We also allude to performance and flexibility facets. A Preliminary implementation and figures from early experimental evaluation are presented.
Current directions in network routing research have not kept pace with the latest developments in... more Current directions in network routing research have not kept pace with the latest developments in network architectures, such as peer-to-peer networks, sensor networks, ad-hoc wireless networks, and overlay networks. A common characteristic among all of these new technologies is the presence of highly dynamic network topologies. Currently deployed single-path routing protocols cannot adequately cope with this dynamism, and existing multi-path algorithms make trade-offs which lead to less than optimal performance on these networks. This drives the need for routing protocols designed with the unique characteristics of these networks in mind. In this paper we propose the notion of reachability routing as a solution to the challenges posed by routing on such dynamic networks. In particular, our formulation of reachability routing provides cost-sensitive multi-path forwarding along with loop avoidance within the confines of the Internet Protocol (IP) architecture. This is achieved throug...
In this paper, we illustrate the impact of dynamic resizability on parallel scheduling. Our ReSHA... more In this paper, we illustrate the impact of dynamic resizability on parallel scheduling. Our ReSHAPE framework includes an application scheduler that supports dynamic resizing of parallel applications. We propose and evaluate new scheduling policies made possible by our ReSHAPE framework. The framework also provides a platform to experiment with more interesting and sophisticated scheduling policies and scenarios for resizable parallel applications. The proposed policies support scheduling of parallel applications with and without user assigned priorities. Experimental results show that these scheduling policies significantly improve individual application turn around time as well as overall cluster utilization.
Grid infrastructures and computing environments have progressed significantly in the past few yea... more Grid infrastructures and computing environments have progressed significantly in the past few years. The vision of truly seamless Grid usage relies on runtime systems support that is cognizant of the operational issues underlying grid computations and, at the same time, is flexible enough to accommodate diverse application scenarios. This paper addresses the twin aspects of Grid infrastructure and application support through a novel combination of two computational technologies – Weaves, a source-language independent parallel runtime compositional framework that operates through reverse-analysis of compiled object files, and runtime recommender systems that aid in dynamic knowledge-based application composition. Domain-specific adaptivity is exploited through a novel compositional system that supports runtime recommendation of code modules and a sophisticated checkpointing and runtime migration solution that can be transparently deployed over Grid infrastructures. A core set of “ada...
Uploads
Papers by Srinidhi Varadarajan