With the rapid evolution of the distributed computing world in the last few years, the amount of data created and processed has fast increased to petabytes or even exabytes scale. Such huge data sets need data-intensive computing applications and impose performance requirements to the infrastructures that support them, such as high scalability, storage, fault tolerance but also efficient scheduling algorithms. This paper focuses on providing a hybrid scheduling algorithm for many task computing that addresses big data environments with few penalties, taking into consideration the deadlines and satisfying a data dependent task model. The hybrid solution consists of several heuristics and algorithms (min-min, min-max and earliest deadline first) combined in order to provide a scheduling algorithm that matches our problem. The experimental results are conducted by simulation and prove that the proposed hybrid algorithm behaves very well in terms of meeting deadlines.
The evolution of ICT systems in the way data is accessed and used is very fast nowadays. Cloud computing is an innovative way of using and providing computing resources to businesses and individuals and it has gained a faster popularity in the last years. In this context, the user’s expectations are increasing and cloud providers are facing huge challenges. One of these challenges is fault tolerance and both researchers and companies have focused on finding and developing strong fault tolerance models. To validate these models, cloud simulation tools are used as an easy, flexible and fast solution. This paper proposes a Fault Injector Module for CloudSim tool (FIM-SIM) for helping the cloud developers to test and validate their infrastructure. FIM-SIM follows the event- driven model and inserts faults in CloudSim based on statistical distributions. The authors have tested and validated it by conducting several experiments designed to highlight the statistical distribution influence on the failures generated and to observe the CloudSim behavior in its current state and implementation.
Context-aware computing is a paradigm that relies on the active use of information coming from a variety of sources, ranging from smartphones to sensors. The paradigm usually leads to storing large volumes of data that need to be processed to derive higher-level context information. The paper presents a cloud-based storage layer for managing sensitive context data. To handle the storage and aggregation of context data for context-aware applications, Clouds are perfect candidates. But a Cloud platform for context-aware computing needs to cope with several requirements: high concurrent access (all data needs to be available to potentially a large number of users), mobility support (such platform should actively use the caches on mobile devices whenever possible, but also cope with storage size limitations), real-time access guarantees – local caches should be located closer to the end-user whenever possible, and persistency (for traceability, a history of the context data should remain available). BlobSeer, a framework for Cloud data storage, is a perfect candidate for storing context data for large-scale applications. It offers capabilities such as persistency, concurrency and support for flexible storage schema requirement. On top of BlobSeer, Context Aware Framework is designed as an extension that offers context-aware data management to higherlevel applications, and enables scalable high-throughput under high-concurrency. On a logical level, the most important capabilities offered by Context Aware Framework are transparency, support for mobility, real-time guarantees and support for access based on meta-information. On the physical layer, the most important capability is persistent Cloud storage.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.