Cloud computing systems are rapidly growing in scale and complexity. They are also changing dynamically as a result of dynamic addition and removal of system components, different execution environments, common updates and upgrades, runtime repairs, mobility of devices and more. Such large-scale, complex and dynamic cloud environments are prone to failures and per- formance anomalies. Thus, dependability and resilience in cloud computing are of paramount importance to guarantee availability and reliability of services and application execution, even in the presence of large number of faulty components. Antifragility is the key to such techniques. It proposes that some systems could be strengthened by changes and faults instead of be weakened by them. In contrast to classical resilience methods, antifragile techniques aim to build systems that handle unpredictable and irregular events, while growing and getting stronger. Most of the classical resilience techniques are not sufficient to build highly available cloud infrastructures. In fact, they just resist shocks and stay the same. They should be complemented by some other aspects like learning from failure to built more elastic and stronger cloud infrastructures. This may represent the idea of building antfragile cloud systems. In this paper, we discuss the existing resilience techniques and propose a solution to design antifragile systems in cloud computing environments

The systematic scaling approach is used in the literature as the only option to meet QoS requirements in response to the traffic load increase in IoT platforms. Such a solution reduces the load on the congested nodes and leads to an increase in the provisioning costs that, when measured at the scale of billions of devices, may hinder acceptability. Our objective is to enhance the existing approaches using and extending the emerging concept of Virtualized Network Functions (VNF), promoting End-to-End IoT traffic control. We start from the observation that, within Cloud-enabled platforms, the allocated resources are not fully used, making it possible to deploy traffic control network functions (NFs). Instead of systematically scaling a congested node, we dynamically deploy, on under-loaded nodes, additional NFs that exploit the available computing resources to differentiate the traffic processing level and to apply a QoS-oriented policy. The considered NFs extend the notion of VNF defined within Network Function Virtualization (NFV), to take advantage of component-based software design. To enable a possible use of the implemented functions, we formulate a multi-objective optimization problem for the efficient planning of adequate NFs and scaling actions, according to the considered multi-constrained context. The planner, called QoS4NIP (QoS for NFV-enabled IoT Platforms), is based on a Genetic Algorithm (GA). The GA metaheuristic relies on biologically inspired operators such as mutation, crossover, and selection. QoS4NIP uses the GA to generate solutions to the identified multi-objective optimization problem by making a series of improvements in an iterative process. The experiments conducted on a realistic case study show that QoS4NIP outperforms autoscaling-based approaches (65% financial cost saved).

Although design patterns have become increasingly popular, most of them are presented in an informal way, which can give rise to ambiguity and may lead to their incorrect usage. Patterns proposed by the SOA design pattern community are described with informal visual notations. Modelling SOA design patterns with a standard formal notation contributes to avoid misunderstanding by software architects and helps endowing design methods with refinement approaches for mastering system architectures complexity. In this paper, we present a formal refinement-based approach that aims, first, to model message-oriented SOA design patterns with the SoaML standard language, and second to formally specify these patterns at a high level of abstraction using the Event-B method. These two steps are performed before undertaking the effective coding of a design pattern providing correct by construction pattern-based software architectures. Our approach is experimented through an example we present in this paper. We implemented our approach under the Rodin platform, which we use to prove model consistency.