HPC management and engineering in the hybrid cloud

The evolution and maturation of Cloud Computing created an opportunity for the emergence of new Cloud applications. High-performance Computing, a complex problem solving class, arises as a new business consumer by taking advantage of the Cloud premises and leaving the expensive datacenter management and difficult grid development. Standing on an advanced maturing phase, today’s Cloud discarded many of its drawbacks, becoming more and more efficient and widespread. Performance enhancements, prices drops due to massification and customizable services on demand triggered an emphasized attention from other markets. HPC, regardless of being a very well established field, traditionally has a narrow frontier concerning its deployment and runs on dedicated datacenters or large grid computing. The problem with common placement is mainly the initial cost and the inability to fully use resources which not all research labs can afford. The main objective of this work was to investigate new technical solutions to allow the deployment of HPC applications on the Cloud, with particular emphasis on the private on-premise resources – the lower end of the chain which reduces costs. The work includes many experiments and analysis to identify obstacles and technology limitations. The feasibility of the objective was tested with new modeling, architecture and several applications migration. The final application integrates a simplified incorporation of both public and private Cloud resources, as well as HPC applications scheduling, deployment and management. It uses a well-defined user role strategy, based on federated authentication and a seamless procedure to daily usage with balanced low cost and performance.

Impact of node mobility in user-centric routing

Recent paradigms in wireless communication architectures describe environments where nodes present a highly dynamic behavior (e.g., User Centric Networks). In such environments, routing is still performed based on the regular packet-switched behavior of store-and-forward. Albeit sufficient to compute at least an adequate path between a source and a destination, such routing behavior cannot adequately sustain the highly nomadic lifestyle that Internet users are today experiencing. This thesis aims to analyse the impact of the nodes’ mobility on routing scenarios. It also aims at the development of forwarding concepts that help in message forwarding across graphs where nodes exhibit human mobility patterns, as is the case of most of the user-centric wireless networks today. The first part of the work involved the analysis of the mobility impact on routing, and we found that node mobility significance can affect routing performance, and it depends on the link length, distance, and mobility patterns of nodes. The study of current mobility parameters showed that they capture mobility partially. The routing protocol robustness to node mobility depends on the routing metric sensitivity to node mobility. As such, mobility-aware routing metrics were devised to increase routing robustness to node mobility. Two categories of routing metrics proposed are the time-based and spatial correlation-based. For the validation of the metrics, several mobility models were used, which include the ones that mimic human mobility patterns. The metrics were implemented using the Network Simulator tool using two widely used multi-hop routing protocols of Optimized Link State Routing (OLSR) and Ad hoc On Demand Distance Vector (AODV). Using the proposed metrics, we reduced the path re-computation frequency compared to the benchmark metric. This means that more stable nodes were used to route data. The time-based routing metrics generally performed well across the different node mobility scenarios used. We also noted a variation on the performance of the metrics, including the benchmark metric, under different mobility models, due to the differences in the node mobility governing rules of the models.