In a wireless sensor network,the sensors cooperate in order to establish the connection between a... more In a wireless sensor network,the sensors cooperate in order to establish the connection between all other sensors in the network.The implementation of this kind of network is done quickly and can be applied in various critical areas,such as disaster relief and military communications.Data Mules Systems,proposed by Shah (2003) has mobile agents called Data Mules (DM),which are used to facilitate the connection between other sensors.The role of these agents is to visit the sensors and perform the delivery and pickup of data between them.Thus,the sensors of the network are classified in 2 ways,simply sensors and special sensors (DM).The sensors not only transmit data in the form of messages to the others or to a base station,but also have an interest in receiving messages from other sensors.The DM are responsible for performing data communication between the sensors using a route that must be defined within the area where sensors are located.The Data Mules Scheduling Problem (DMSP) is divided in 3 phases.The phase of Path Selection consists on defining a route to collect and delivery information from each sensor.In Speed Control phase,the management of speed of the Data Mule through the route is defined.And finally,the Job Scheduling phase is important when the mule is in contact with more than one sensor at the same time,so since the DM can change data with only one sensor at a time,this phase is responsible for the scheduling of communications.The goal of DMSP is serve the demand of all sensors by minimizing the time spend in the route.In many works in the literature,the DMSP is divided into two stages,the first one being the definition of the path of DM and the second stage deal with speed and scheduling problems.The second stage is only started when the first stage is complete.This approach isn’t advantageous because the decision of a stage can influence the other one,so in this work we propose a mathematical formulation to represent the problem in a unified way,using adapted instances for the Close-Enough TSP to validate the model
Anais do XVIII Simpósio em Sistemas Computacionais de Alto Desempenho (WSCAD 2017)
GPUs have established a new baseline for power efficiency and computing power, delivering larger ... more GPUs have established a new baseline for power efficiency and computing power, delivering larger bandwidth and more computing units in each new generation. Modern GPUs support the concurrent execution of kernels to maximize resource utilization, allowing other kernels to better exploit idle resources. However, the decision on the simultaneous execution of different kernels is made by the hardware, and sometimes GPUs do not allow the execution of blocks from other kernels, even with the availability of resources. In this work, we present an in-depth study on the simultaneous execution of kernels on the GPU. We present the necessary conditions for executing kernels simultaneously, we define the factors that influence competition, and describe a model that can determine performance degradation. Finally, we validate the model using synthetic and real-world kernels with different computation and memory requirements.
The Pre-stack Kirchhoff Time Migration (PKTM) is a central process in petroleum exploration. As P... more The Pre-stack Kirchhoff Time Migration (PKTM) is a central process in petroleum exploration. As PKTM is computationally intensive, many works have proposed the use of accelerators like GPU and FPGA to improve its execution time. On the other hand, although many off-the-shelf processors are endowed with a set of SIMD vector instructions, few papers tackle the problem considering vectorization and, all of them, consider that compilers can successfully vectorize the code. In this paper, we show that a handwritten Kirchhoff code by using SIMD vector instructions is more efficient than the autovectorized code provided by GCC. Experiments considering both real and synthetic datasets showed that our solution is up to eight times faster than the autovectorized one. We believe that the proposed strategy can be used together with the other ones to accelerate seismic migration methods in general without new investments in hardware and software.
Anais do XXI Workshop de Testes e Tolerância a Falhas (WTF 2020), 2020
Os provedores de nuvem oferecem vários serviços para processamento e armazenamento de dados. Algu... more Os provedores de nuvem oferecem vários serviços para processamento e armazenamento de dados. Alguns desses recursos são propensos a falhas ou revogações. Neste caso, a adoção de técnicas de tolerância a falhas é essencial para minimizar o impacto e o custo que estas falhas podem exercer na execução das aplicações e no orçamento dos usuários. Uma das soluções amplamente adotadas são as técnicas de checkpoint and recovery (CR), que gravam o estado atual da aplicação para que, caso ocorra falhas, esta seja reiniciada a partir do último estado gravado. Assim, essas abordagens exigem um local de armazenamento seguro e disponível ao longo de toda a execução. Felizmente, os serviços de nuvens também podem ser utilizados para este fim, uma vez que diferentes opções de armazenamento também são oferecidos. Neste trabalho, queremos avaliar três desses serviços: o Amazon Simple Storage Service (S3), o Amazon Elastic Block Store (EBS) e o Amazon Elastic File System (EFS). O objetivo é caracteriz...
Leading Cloud providers offer several types of Virtual Machines (VMs) in diverse contract models,... more Leading Cloud providers offer several types of Virtual Machines (VMs) in diverse contract models, with different guarantees in terms of availability and reliability. Among them, the most popular contract models are the on-demand and the spot models. In the former, on-demand VMs are allocated for a fixed cost per time unit, and their availability is ensured during the whole execution. On the other hand, in the spot market, VMs are offered with a huge discount when compared to the ondemand VMs, but their availability fluctuates according to the cloud's current demand that can terminate or hibernate a spot VM at any time. Furthermore, in order to cope with workload variations, cloud providers have also introduced the concept of burstable VMs which are able to burst up their respective baseline CPU performance during a limited period of time with an up to 20% discount when compared to an equivalent nonburstable on-demand VMs. In the current work, we present the Burst Hibernation-Aware Dynamic Scheduler (Burst-HADS), a framework that schedules and executes tasks of Bag-of-Tasks applications with deadline constraints by exploiting spot and ondemand burstable VMs, aiming at minimizing both the monetary cost and the execution time. Based on ILS metaheuristics, Burst-HADS defines an initial scheduling map of tasks to VMs which can then be dynamically altered by migrating tasks of a hibernated spot VM or by performing work-stealing when VMs become idle. Performance results on Amazon EC2 cloud with different applications show that, when compared to a solution that uses only regular on-demand instances, Burst-HADS reduces the monetary cost of the execution and meet the application deadline even in scenarios with high spot hibernation rates. It also reduces the total execution time when compared to a solution that uses only spot and non-burstable on-demand instances.
Matrix clocks are a generalization of the notion of vector clocks that allows the local represent... more Matrix clocks are a generalization of the notion of vector clocks that allows the local representation of causal precedence to reach into an asynchronous distributed computation's past with depth x, where x ≥ 1 is an integer. Maintaining matrix clocks correctly in a system of n nodes requires that every message be accompanied by O(n x ) numbers, which reflects an exponential dependency of the complexity of matrix clocks upon the desired depth x. We introduce a novel type of matrix clock, one that requires only nx numbers to be attached to each message while maintaining what for many applications may be the most significant portion of the information that the original matrix clock carries. In order to illustrate the new clock's applicability, we demonstrate its use in the monitoring of certain resource-sharing computations.
2019 31st International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), Oct 1, 2019
Cloud platforms offer different types of virtual machines which ensure different guarantees in te... more Cloud platforms offer different types of virtual machines which ensure different guarantees in terms of availability and volatility, provisioning the same resource through multiple pricing models. For instance, in Amazon EC2 cloud, the user pays per hour for on-demand instances while spot instances are unused resources available for a lower price. Despite the monetary advantages, a spot instance can be terminated or hibernated by EC2 at any moment. Using both hibernationprone spot instances (for cost sake) and on-demand instances, we propose in this paper a static scheduling for applications which are composed of independent tasks (bag-of-task) with deadline constraints. However, if a spot instance hibernates and it does not resume within a time which guarantees the application's deadline, a temporal failure takes place. Our scheduling, thus, aims at minimizing monetary costs of bag-of-tasks applications in EC2 cloud, respecting its deadline and avoiding temporal failures. Performance results with task execution traces, configuration of Amazon EC2 virtual machines, and EC2 market history confirms the effectiveness of our scheduling and that it tolerates temporal failures.
Data mining techniques are becoming increasingly more popular as reasonable means to collect summ... more Data mining techniques are becoming increasingly more popular as reasonable means to collect summaries from the rapidly growing datasets in many areas. However, as the size of the raw data increases, parallel data mining algorithms are becoming a necessity. In this paper we present a run-time support system that was designed to allow the efficient implementation of data-mining algorithms on heterogeneous distributed environments. We believe that the runtime framework is suitable for a broader class of applications, beyond data mining. We also present, through an example, a parallelization strategy that is supported by the run-time system. We show scalability results of two different data-mining algorithms that were parallelized using our approach and our runt-time support. Both applications scale very close to linearly with large number of nodes.
This paper presents an asynchronous parallel metaheuristic for the period vehicle routing problem... more This paper presents an asynchronous parallel metaheuristic for the period vehicle routing problem (PVRP). The PVRP generalizes the classical vehicle routing problem by extending the planning period from a single day to M days. The algorithm proposed is based on concepts used in parallel genetic algorithms and local search heuristics. The algorithm employs the Island model in which the migration frequency must not be very high. The results of computational experiments carried out on problems taken from the literature indicate that the proposed approach outperforms existing heuristics in most cases.
Nowadays genetic algorithms stand as a trend to solve NPcomplete and NP-hard problems. In this pa... more Nowadays genetic algorithms stand as a trend to solve NPcomplete and NP-hard problems. In this paper, we present a new hybrid metaheuristic which uses Parallel Genetic Algorithms and Scatter Search coupled with a decomposition-into-petals procedure for solving a class of Vehicle Routing and Scheduling Problems. The parallel genetic algorithm presented is based on the island model and was run on a cluster of workstations. Its performance is evaluated for a heterogeneous fleet problem, which is considered a problem much harder to solve than the homogeneous vehicle routing problem.
This paper proposes the Makespan and Reliability Cost Driven (MRCD) heuristic, a static schedulin... more This paper proposes the Makespan and Reliability Cost Driven (MRCD) heuristic, a static scheduling strategy for heterogeneous distributed systems that not only minimizes the makespan, but also maximizes the reliability of the application. The MRCD scheduling decisions are guided by a weighted function that considers both objectives simultaneously, instead of prioritizing one of them. This work also introduces a classification of the solutions produced by weighted bi-objective schedulers to aid users to tune the weighting function such that an appropriate solution can be selected in accordance with their needs. In comparison with the related work, MRCD produced schedules with makespans that were significantly better then those produced by the other strategies at expense of an insignificant deterioration in reliability.► A heterogeneous system scheduler minimizes makespan and maximizes reliability. ► The static scheduling decisions are guided by a weighted bi-objective function. ► The classification of scheduler solutions aids users to tune the weighted function. ► Scheduler solutions were compared with up-to-dated related work. ► Makespans were improved while an insignificant deterioration in reliability occurred.
* Department of Computer Science, Fluminense Federal University Rua Passo da Patria, 156, Bloco E... more * Department of Computer Science, Fluminense Federal University Rua Passo da Patria, 156, Bloco E, Niteroi, Rio de Janeiro, 24210-240, BRAZIL Email: {lucia, satoru}@dcc.ic.uff.br ... Metaheuristics such as Greedy Randomized Adaptive Search Procedures ...
Content Distribution Networks (CDN) are overlay networks used to place content near end clients w... more Content Distribution Networks (CDN) are overlay networks used to place content near end clients with the aim at reducing delay, servers load and network congestion, hence improving the service quality perceived by end clients.
Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados através da ... more Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados através da internet que colaboram para fornecer conteúdos para clientes. Ela é uma rede sobreposta que mantém réplicas de cada conteúdo em seus servidores, com o objetivo de reduzir atrasos, a carga nos servidores e o congestionamento da rede, melhorando a qualidade do serviço. Em arquiteturas de RDC tradicionais, as requisições são recebidas por um servidor central e redirecionadas para o servidor mais próximo do cliente. Entretanto, devido aos custos de manutenção das réplicas, não é razoável replicar os conteúdos em toda a rede. Neste trabalho são apresentadas abordagens exatas e heurísticas para resolver um problema de gerenciamento das RDC, o Problema de Posicionamento de Réplicas (PPR). Este problema consiste em decidir quais réplicas serão colocadas em cada servidor. O objetivo final é minimizar o tráfego na rede sem violar as restrições de qualidade de serviço (QoS).
Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados que colabor... more Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados que colaboram para fornecer conteúdos para clientes. Ela mantém réplicas de cada conteúdo em seus servidores, objetivando reduzir atrasos, a carga nos servidores e o congestionamento da rede, melhorando a qualidade do serviço. Em arquiteturas de RDC tradicionais, as requisições são recebidas por um servidor central e redirecionadas para o servidor mais próximo do cliente. Entretanto, devido aos custos de replicação, não é razoável replicar os conteúdos em toda a rede. Neste trabalho são apresentadas abordagens exatas e heurísticas para resolver um problema de gerenciamento das RDC, o Problema de Posicionamento de Réplicas e Distribuição de Requisições (PPRDR) online. Este problema consiste em decidir quais réplicas serão colocadas em cada servidor e quais servidores irão atender as requisições. O objetivo final é minimizar o tráfego na rede sem violar as restrições de qualidade de serviço (QoS).
In a wireless sensor network,the sensors cooperate in order to establish the connection between a... more In a wireless sensor network,the sensors cooperate in order to establish the connection between all other sensors in the network.The implementation of this kind of network is done quickly and can be applied in various critical areas,such as disaster relief and military communications.Data Mules Systems,proposed by Shah (2003) has mobile agents called Data Mules (DM),which are used to facilitate the connection between other sensors.The role of these agents is to visit the sensors and perform the delivery and pickup of data between them.Thus,the sensors of the network are classified in 2 ways,simply sensors and special sensors (DM).The sensors not only transmit data in the form of messages to the others or to a base station,but also have an interest in receiving messages from other sensors.The DM are responsible for performing data communication between the sensors using a route that must be defined within the area where sensors are located.The Data Mules Scheduling Problem (DMSP) is divided in 3 phases.The phase of Path Selection consists on defining a route to collect and delivery information from each sensor.In Speed Control phase,the management of speed of the Data Mule through the route is defined.And finally,the Job Scheduling phase is important when the mule is in contact with more than one sensor at the same time,so since the DM can change data with only one sensor at a time,this phase is responsible for the scheduling of communications.The goal of DMSP is serve the demand of all sensors by minimizing the time spend in the route.In many works in the literature,the DMSP is divided into two stages,the first one being the definition of the path of DM and the second stage deal with speed and scheduling problems.The second stage is only started when the first stage is complete.This approach isn’t advantageous because the decision of a stage can influence the other one,so in this work we propose a mathematical formulation to represent the problem in a unified way,using adapted instances for the Close-Enough TSP to validate the model
Anais do XVIII Simpósio em Sistemas Computacionais de Alto Desempenho (WSCAD 2017)
GPUs have established a new baseline for power efficiency and computing power, delivering larger ... more GPUs have established a new baseline for power efficiency and computing power, delivering larger bandwidth and more computing units in each new generation. Modern GPUs support the concurrent execution of kernels to maximize resource utilization, allowing other kernels to better exploit idle resources. However, the decision on the simultaneous execution of different kernels is made by the hardware, and sometimes GPUs do not allow the execution of blocks from other kernels, even with the availability of resources. In this work, we present an in-depth study on the simultaneous execution of kernels on the GPU. We present the necessary conditions for executing kernels simultaneously, we define the factors that influence competition, and describe a model that can determine performance degradation. Finally, we validate the model using synthetic and real-world kernels with different computation and memory requirements.
The Pre-stack Kirchhoff Time Migration (PKTM) is a central process in petroleum exploration. As P... more The Pre-stack Kirchhoff Time Migration (PKTM) is a central process in petroleum exploration. As PKTM is computationally intensive, many works have proposed the use of accelerators like GPU and FPGA to improve its execution time. On the other hand, although many off-the-shelf processors are endowed with a set of SIMD vector instructions, few papers tackle the problem considering vectorization and, all of them, consider that compilers can successfully vectorize the code. In this paper, we show that a handwritten Kirchhoff code by using SIMD vector instructions is more efficient than the autovectorized code provided by GCC. Experiments considering both real and synthetic datasets showed that our solution is up to eight times faster than the autovectorized one. We believe that the proposed strategy can be used together with the other ones to accelerate seismic migration methods in general without new investments in hardware and software.
Anais do XXI Workshop de Testes e Tolerância a Falhas (WTF 2020), 2020
Os provedores de nuvem oferecem vários serviços para processamento e armazenamento de dados. Algu... more Os provedores de nuvem oferecem vários serviços para processamento e armazenamento de dados. Alguns desses recursos são propensos a falhas ou revogações. Neste caso, a adoção de técnicas de tolerância a falhas é essencial para minimizar o impacto e o custo que estas falhas podem exercer na execução das aplicações e no orçamento dos usuários. Uma das soluções amplamente adotadas são as técnicas de checkpoint and recovery (CR), que gravam o estado atual da aplicação para que, caso ocorra falhas, esta seja reiniciada a partir do último estado gravado. Assim, essas abordagens exigem um local de armazenamento seguro e disponível ao longo de toda a execução. Felizmente, os serviços de nuvens também podem ser utilizados para este fim, uma vez que diferentes opções de armazenamento também são oferecidos. Neste trabalho, queremos avaliar três desses serviços: o Amazon Simple Storage Service (S3), o Amazon Elastic Block Store (EBS) e o Amazon Elastic File System (EFS). O objetivo é caracteriz...
Leading Cloud providers offer several types of Virtual Machines (VMs) in diverse contract models,... more Leading Cloud providers offer several types of Virtual Machines (VMs) in diverse contract models, with different guarantees in terms of availability and reliability. Among them, the most popular contract models are the on-demand and the spot models. In the former, on-demand VMs are allocated for a fixed cost per time unit, and their availability is ensured during the whole execution. On the other hand, in the spot market, VMs are offered with a huge discount when compared to the ondemand VMs, but their availability fluctuates according to the cloud's current demand that can terminate or hibernate a spot VM at any time. Furthermore, in order to cope with workload variations, cloud providers have also introduced the concept of burstable VMs which are able to burst up their respective baseline CPU performance during a limited period of time with an up to 20% discount when compared to an equivalent nonburstable on-demand VMs. In the current work, we present the Burst Hibernation-Aware Dynamic Scheduler (Burst-HADS), a framework that schedules and executes tasks of Bag-of-Tasks applications with deadline constraints by exploiting spot and ondemand burstable VMs, aiming at minimizing both the monetary cost and the execution time. Based on ILS metaheuristics, Burst-HADS defines an initial scheduling map of tasks to VMs which can then be dynamically altered by migrating tasks of a hibernated spot VM or by performing work-stealing when VMs become idle. Performance results on Amazon EC2 cloud with different applications show that, when compared to a solution that uses only regular on-demand instances, Burst-HADS reduces the monetary cost of the execution and meet the application deadline even in scenarios with high spot hibernation rates. It also reduces the total execution time when compared to a solution that uses only spot and non-burstable on-demand instances.
Matrix clocks are a generalization of the notion of vector clocks that allows the local represent... more Matrix clocks are a generalization of the notion of vector clocks that allows the local representation of causal precedence to reach into an asynchronous distributed computation's past with depth x, where x ≥ 1 is an integer. Maintaining matrix clocks correctly in a system of n nodes requires that every message be accompanied by O(n x ) numbers, which reflects an exponential dependency of the complexity of matrix clocks upon the desired depth x. We introduce a novel type of matrix clock, one that requires only nx numbers to be attached to each message while maintaining what for many applications may be the most significant portion of the information that the original matrix clock carries. In order to illustrate the new clock's applicability, we demonstrate its use in the monitoring of certain resource-sharing computations.
2019 31st International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), Oct 1, 2019
Cloud platforms offer different types of virtual machines which ensure different guarantees in te... more Cloud platforms offer different types of virtual machines which ensure different guarantees in terms of availability and volatility, provisioning the same resource through multiple pricing models. For instance, in Amazon EC2 cloud, the user pays per hour for on-demand instances while spot instances are unused resources available for a lower price. Despite the monetary advantages, a spot instance can be terminated or hibernated by EC2 at any moment. Using both hibernationprone spot instances (for cost sake) and on-demand instances, we propose in this paper a static scheduling for applications which are composed of independent tasks (bag-of-task) with deadline constraints. However, if a spot instance hibernates and it does not resume within a time which guarantees the application's deadline, a temporal failure takes place. Our scheduling, thus, aims at minimizing monetary costs of bag-of-tasks applications in EC2 cloud, respecting its deadline and avoiding temporal failures. Performance results with task execution traces, configuration of Amazon EC2 virtual machines, and EC2 market history confirms the effectiveness of our scheduling and that it tolerates temporal failures.
Data mining techniques are becoming increasingly more popular as reasonable means to collect summ... more Data mining techniques are becoming increasingly more popular as reasonable means to collect summaries from the rapidly growing datasets in many areas. However, as the size of the raw data increases, parallel data mining algorithms are becoming a necessity. In this paper we present a run-time support system that was designed to allow the efficient implementation of data-mining algorithms on heterogeneous distributed environments. We believe that the runtime framework is suitable for a broader class of applications, beyond data mining. We also present, through an example, a parallelization strategy that is supported by the run-time system. We show scalability results of two different data-mining algorithms that were parallelized using our approach and our runt-time support. Both applications scale very close to linearly with large number of nodes.
This paper presents an asynchronous parallel metaheuristic for the period vehicle routing problem... more This paper presents an asynchronous parallel metaheuristic for the period vehicle routing problem (PVRP). The PVRP generalizes the classical vehicle routing problem by extending the planning period from a single day to M days. The algorithm proposed is based on concepts used in parallel genetic algorithms and local search heuristics. The algorithm employs the Island model in which the migration frequency must not be very high. The results of computational experiments carried out on problems taken from the literature indicate that the proposed approach outperforms existing heuristics in most cases.
Nowadays genetic algorithms stand as a trend to solve NPcomplete and NP-hard problems. In this pa... more Nowadays genetic algorithms stand as a trend to solve NPcomplete and NP-hard problems. In this paper, we present a new hybrid metaheuristic which uses Parallel Genetic Algorithms and Scatter Search coupled with a decomposition-into-petals procedure for solving a class of Vehicle Routing and Scheduling Problems. The parallel genetic algorithm presented is based on the island model and was run on a cluster of workstations. Its performance is evaluated for a heterogeneous fleet problem, which is considered a problem much harder to solve than the homogeneous vehicle routing problem.
This paper proposes the Makespan and Reliability Cost Driven (MRCD) heuristic, a static schedulin... more This paper proposes the Makespan and Reliability Cost Driven (MRCD) heuristic, a static scheduling strategy for heterogeneous distributed systems that not only minimizes the makespan, but also maximizes the reliability of the application. The MRCD scheduling decisions are guided by a weighted function that considers both objectives simultaneously, instead of prioritizing one of them. This work also introduces a classification of the solutions produced by weighted bi-objective schedulers to aid users to tune the weighting function such that an appropriate solution can be selected in accordance with their needs. In comparison with the related work, MRCD produced schedules with makespans that were significantly better then those produced by the other strategies at expense of an insignificant deterioration in reliability.► A heterogeneous system scheduler minimizes makespan and maximizes reliability. ► The static scheduling decisions are guided by a weighted bi-objective function. ► The classification of scheduler solutions aids users to tune the weighted function. ► Scheduler solutions were compared with up-to-dated related work. ► Makespans were improved while an insignificant deterioration in reliability occurred.
* Department of Computer Science, Fluminense Federal University Rua Passo da Patria, 156, Bloco E... more * Department of Computer Science, Fluminense Federal University Rua Passo da Patria, 156, Bloco E, Niteroi, Rio de Janeiro, 24210-240, BRAZIL Email: {lucia, satoru}@dcc.ic.uff.br ... Metaheuristics such as Greedy Randomized Adaptive Search Procedures ...
Content Distribution Networks (CDN) are overlay networks used to place content near end clients w... more Content Distribution Networks (CDN) are overlay networks used to place content near end clients with the aim at reducing delay, servers load and network congestion, hence improving the service quality perceived by end clients.
Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados através da ... more Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados através da internet que colaboram para fornecer conteúdos para clientes. Ela é uma rede sobreposta que mantém réplicas de cada conteúdo em seus servidores, com o objetivo de reduzir atrasos, a carga nos servidores e o congestionamento da rede, melhorando a qualidade do serviço. Em arquiteturas de RDC tradicionais, as requisições são recebidas por um servidor central e redirecionadas para o servidor mais próximo do cliente. Entretanto, devido aos custos de manutenção das réplicas, não é razoável replicar os conteúdos em toda a rede. Neste trabalho são apresentadas abordagens exatas e heurísticas para resolver um problema de gerenciamento das RDC, o Problema de Posicionamento de Réplicas (PPR). Este problema consiste em decidir quais réplicas serão colocadas em cada servidor. O objetivo final é minimizar o tráfego na rede sem violar as restrições de qualidade de serviço (QoS).
Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados que colabor... more Uma Rede de Distribuição de Conteúdos (RDC) é um sistema de computadores interligados que colaboram para fornecer conteúdos para clientes. Ela mantém réplicas de cada conteúdo em seus servidores, objetivando reduzir atrasos, a carga nos servidores e o congestionamento da rede, melhorando a qualidade do serviço. Em arquiteturas de RDC tradicionais, as requisições são recebidas por um servidor central e redirecionadas para o servidor mais próximo do cliente. Entretanto, devido aos custos de replicação, não é razoável replicar os conteúdos em toda a rede. Neste trabalho são apresentadas abordagens exatas e heurísticas para resolver um problema de gerenciamento das RDC, o Problema de Posicionamento de Réplicas e Distribuição de Requisições (PPRDR) online. Este problema consiste em decidir quais réplicas serão colocadas em cada servidor e quais servidores irão atender as requisições. O objetivo final é minimizar o tráfego na rede sem violar as restrições de qualidade de serviço (QoS).
Uploads
Papers by lucia drummond