As the popularity of cloud computing increases, more and more applications are migrated onto them. Web 2.0 applications are the most common example of such applications. These applications require to scale, be highly available, fault tolerant and able to run uninterrupted for long periods of time (or even indefinitely). Moreover as new cloud providers appear there is a natural tendency towards choosing the best provider or a combination of them for deploying the application. Thus multi-cloud scenarios emerge from this situation. However, as multi-cloud resource provisioning is both complex and costly, the choice of which resources to lend and how to allocate them to application components needs to rely on efficient strategies. These need to take into account many factors including deployment and run-time cost, resource load, and application availability in case of failures. For this aim multi-objective scheduling algorithms seem an appropriate choice. This paper presents an algorithm which tries to achieve application high-availability and fault-tolerance while reducing the application cost and keeping the resource load maximized. The proposed algorithm is compared with a classic Round Robin strategy - used by many commercial clouds - and the obtained results prove the efficiency of our solution.

Current Platform as a Service solutions restrict the user to certain programming languages, paradigms or Cloud service types. A new deployable open-source Platform as a Service that includes a set of extensible APIs eases the deployment and control of the software stack required by Cloud applications. This paper describes shortly its vision, concepts and status of implementation with focus on a particular platform component used for controlling the deployment and behavior of the application's components.

The increased volume of images and galaxies surveyed by recent and upcoming projects consolidates the need for accurate and scalable automated AI-driven classification methods. This paper proposes a new algorithm based on a custom neural network architecture for classifying galaxies from deep space surveys. The convolutional neural network (CNN) presented is trained using 10,000 galaxy images obtained from the Galaxy Zoo 2 dataset. It is designed to categorize galaxies into five distinct classes: completely round smooth, in-between smooth (falling between completely round and cigar-shaped), cigar-shaped smooth, edge-on, and spiral. The performance of the proposed CNN is assessed using a set of metrics such as accuracy, precision, recall, F1 score, and area under the curve. We compare our solution with well-known architectures like ResNet-50, DenseNet, EfficientNet, Inception, MobileNet, and one proposed model for galaxy classification found in the recent literature. The results show an accuracy rate of 96.83%, outperforming existing algorithms.

Distributed systems require effective mechanisms to manage the reliable provisioning of computational resources from different and distributed providers. Moreover, the dynamic environment that affects the behaviour of such systems and the complexity of these dynamics demand autonomous capabilities to ensure the behaviour of distributed scheduling platforms and to achieve business and user objectives. In this paper we propose a self-adaptive distributed scheduling platform composed of multiple agents implemented as intelligent feedback control loops to support policy-based scheduling and expose self-healing capabilities. Our platform leverages distributed scheduling processes by (i) allowing each provider to maintain its own internal scheduling process, and (ii) implementing self-healing capabilities based on agent module recovery. Simulated tests are performed to determine the optimal number of agents to be used in the negotiation phase without affecting the scheduling cost function. Test results on a real-life platform are presented to evaluate recovery times and optimize platform parameters.