Multiple sequence alignment (MSA) is one of the most useful tools in bioinformatics. MSA plays a key role in protein/RNA structure prediction, phylogenetic analysis or pattern identification among other important bioinformatic applications. However, the growth of sequencing data imposes further difficulties to aligning it with traditional tools. For large-scale alignments with thousands of sequences or even whole genomes, it will be necessary to use and take advantage of high performance computing (HPC). This paper, focused on a consistency-based MSA tool called T-Coffee, presents several innovative solutions; the Balanced Guide Tree, the Optimized Library Method and the Multiple Tree Alignment. The results obtained by the methods presented have demonstrated that it is possible to improve efficiency, scalability and accuracy.

No abstract

In order to ensure a more widespread implementation of video-ondemand (VoD) services, it is essential that the design of cost-effective largescale VoD (LVoD) architectures be able to support hundreds of thousands of concurrent users. The main keys for the designing of such architectures are high streaming capacity, low costs, scalability, fault tolerance, load balance, low complexity and resource sharing among user requests. To achieve these objectives, we propose a distributed architecture, called double P-Tree, which is based on a tree topology of independent local networks with proxies. The proxy functionality has been modified in such a way that it works at the same time as cache for the most-watched videos, and as a distributed mirror for the remaining videos. In this way, we manage to distribute main server functionality (as a repository of all system videos, server of proxy-misses and system manager) among all local proxies. The evaluation of this new architecture, through an analytical model, shows that double P-Tree architecture is a good approach for the building of scalable and fault-tolerant LVoD systems. Experimental results show that this architecture achieves a good tradeoff between effective bandwidth and storage requirements.

No abstract

No abstract

No abstract

Multi-cluster environments are composed of multiple clusters of computers that act collaboratively, and thus allowing computational problems to be treated that require more resources than those available in a single cluster. However, the degree of complexity of the scheduling process is greatly increased by the heterogeneity of resources and co-allocation process, which distributes the tasks of parallel jobs across cluster boundaries.This work presents a new scheduling strategy that allocates multiple jobs from the system queue simultaneously on a heterogeneous multicluster, by applying coallocation when is necessary. Our strategy is composed by a job selection function and a linear programming model to find the best allocation for multiple jobs. The proposed scheduling technique is shown to reduce the execution times of the parallel jobs and the overall response times by 38% compared with other scheduling techniques in the literature.

The advance of Internet 2 and the proliferation of switches and routers with level three functionalities made the multicast one of the most feasible video streaming delivering techniques for the near future. Assuming this to be true, this study addressed the overload situation that a streaming server could suffer due to client requests. As a solution, we proposed new multicast delivery scheme that allows every active client to collaborate with the server regardless of the video that they are watching, alleviating server loads, and therefore server resource requirements. The solution combined the multicast delivery scheme and client-side buffer collaboration in order to decentralize the delivery process. The new video delivering scheme was designed as two separate policies: the first policy used client collaboration to deliver first part of videos and the second policy could merge two or more multicast channels using distributed collaboration between a group of clients. Experimental results show that this scheme is better than previous schemes in terms of resource requirements and scalability.