Danilo Segura

In recent years, cloud computing has been one of the most widely discussed topics in the field of Information Technology. Owing to the popularity of services offered by cloud environments, several critical aspects of security have aroused interest in the academic and industrial world, where there is a concern to provide efficient mechanisms to combat a wide range of threats. As is well known, the application of security techniques and methodologies has a direct influence on the performance of… Mostrar más

In recent years, cloud computing has been one of the most widely discussed topics in the field of Information Technology. Owing to the popularity of services offered by cloud environments, several critical aspects of security have aroused interest in the academic and industrial world, where there is a concern to provide efficient mechanisms to combat a wide range of threats. As is well known, the application of security techniques and methodologies has a direct influence on the performance of the system, since security and performance are two quantities that are inversely proportional. This means that if the service providers fail to manage their computing infrastructure efficiently, the demand for services may not be met with the quality required by clients, including security and performance requirements, and the computational resources may be used inefficiently. The aim of this paper was to define QoS-driven approaches for cloud environments on the basis of the results of a performance evaluation of a service in which different security mechanisms are employed. These mechanisms impose additional overhead on the performance of the service, and to counter this, an attempt was made to change computational resources dynamically and on-the-fly. On the basis of the results, it could be shown that in a cloud environment, it is possible to maintain the performance of the service even with the overhead imposed by the security mechanisms, through an alteration in the virtualized computational resources. However, this change in the amount of resources had a direct effect on the response variables. Mostrar menos

Abstract:
This paper presents an approach to provide quality of service for IoT-Cloud s through a dynamic virtual machine allocation algorithm on a Metascheduler. Although there are others reactive approach to handle with general cases of virtual machine allocation and placement, they may have not optimal performance for IoT workload, because it aggregates a new characteristic of large bursts of events, but with short service time. For this, in this paper, a predictive dynamic algorithm… Mostrar más

Abstract:
This paper presents an approach to provide quality of service for IoT-Cloud s through a dynamic virtual machine allocation algorithm on a Metascheduler. Although there are others reactive approach to handle with general cases of virtual machine allocation and placement, they may have not optimal performance for IoT workload, because it aggregates a new characteristic of large bursts of events, but with short service time. For this, in this paper, a predictive dynamic algorithm from our Metascheduler is outlined which can provide QoS. Our approach combines several informations to build a operating virtual machines to perform the services from IoT gateways devices deployed in the field, that conform to the constraints imposed by QoS with regards to costs and deadlines. As a part of the results obtained from the planning design, our experiments on evaluation performance show the value of this algorithm in meeting the requirements time e cost to the IoT-s, while also saving money and computer resources. Mostrar menos

Abstract:
The goals of a Distributed File Systems (DFS) may vary broadly. It is impossible to design a DFS attaining every desirable characteristic, such as, transparency, performance, privacy, reliability, and availability, for example. In this paper we describe the improvements achieved with the availability and performance offered by a DFS named FlexA (Flexible and Adaptable Distributed File System), which already proposed an architecture that could provide data security and flexibility… Mostrar más

Abstract:
The goals of a Distributed File Systems (DFS) may vary broadly. It is impossible to design a DFS attaining every desirable characteristic, such as, transparency, performance, privacy, reliability, and availability, for example. In this paper we describe the improvements achieved with the availability and performance offered by a DFS named FlexA (Flexible and Adaptable Distributed File System), which already proposed an architecture that could provide data security and flexibility. Modifications included a new approach to provide file replication and procedures to prevent system overloads. Details about the modifications introduced to FlexA, as well as results achieved with them, are provided. These results indicate that FlexA can be an important option among the known DFS. Mostrar menos