Cloud computing systems can benefit from the use of personal and non-dedicated computers, which are currently employed in volunteer computing systems. Being non-dedicated, these resources show random behavior regarding the times they are online (available) and offline. Accordingly, their availability levels are lower than those of traditionally employed dedicated resources. Thus, in order to use non-dedicated resources in cloud computing environments, it is necessary first to solve the problem of how to attain high availability levels for the Internet services deployed over them. Most approaches on how to guarantee high service availability levels with non-dedicated resources are based on the introduction of high degrees of redundancy into the system. However, this praxis leads to an inefficient usage of computational resources and, therefore, to higher operational costs. Accordingly, the focus of this paper is the problem of minimizing the cost of a service deployment over non-dedicated resources while providing a high level of service availability. In order to solve this stochastic optimization problem, the paper proposes a hybrid algorithm that combines a metaheuristic component with a discrete-event simulation component. The metaheuristic component is used to search for an efficient configuration of resources. The simulation component is integrated inside the metaheuristic and used to estimate the service availability of each promising configuration. A numerical experiment section, comparing the performance of several algorithms, contributes to validating the proposed approach as well as to illustrate its potential applications.

Community Cloud computing is a new trend on cloud computing that aims to build service infrastructures upon Wireless Community Networks taking advantage of underused community physical resources. Service allocation protocols are a key design challenge that all cloud systems must properly address to optimize resource utilization. They are specially important when cloud services require a Quality of Service (QoS) and network stability or performance (delay, jitter, minimum bandwidth) cannot be guaranteed a-priory. This work presents a study that tries to understand how to address cloud service deployments in such scenario. In particular, we start proposing an allocation algorithm to find optimal solutions when there is a central authority that coordinates the process. These solutions optimize the communication cost in two ways: (1) minimizing the service overlay diameter and, (2) minimizing the coordination cost along the network. Based on the study of the algorithm and the experimental simulations, we study the variables that outcome optimal service allocations to the detriment of other solutions. We verify these findings using data mining techniques. Researchers can take advantage of the simulation results and our observations to design more reliable distributed algorithms able to dynamically self-adapt to network changes.

This paper introduces a probabilistic algorithm for solving the well-known Facility Location Problem (FLP), an optimization problem frequently encountered in practical applications in fields such as Logistics or Telecommunications. Our algorithm is based on the combination of biased random sampling -using a skewed probability distribution-with a metaheuristic framework. The use of random variates from a skewed distribution allows to guide the local search process inside the metaheuristic framework which, being a stochastic procedure, is likely to produce slightly different results each time it is run. Our approach is validated against some classical benchmarks from the FLP literature and it is also used to analyze the deployment of service replicas in a realistic Internet-distributed system.

No abstract