Learn More
We introduce a new dataset, Human3.6M, of 3.6 Million accurate 3D Human poses, acquired by recording the performance of 5 female and 6 male subjects, under 4 different viewpoints, for training realistic human sensing systems and for evaluating the next generation of human pose estimation models and algorithms. Besides increasing the size of the datasets in(More)
This paper presents the integration of two collective I/O techniques into the Clusterfile parallel file system: disk-directed I/O and two-phase I/O. We show that global cooperative cache management improves the collective I/O performance. The solution focuses on integrating disk parallelism with other types of parallelism: memory (by buffering and caching(More)
This work presents a performance analysis of request distribution-aware caching in cluster-based Web servers. We use the Zipf-like request distribution curve to guide static Web document caching. A combination of cooperative caching and exclusive caching provides for a cluster-wide caching system that avoids document replication accross the cluster. We(More)
BACKGROUND The incidence of urologic complications after kidney transplantation remains high despite improvements in diagnosis and operative techniques. Urinary tract reconstruction is usually done by ureteroneocystostomy (UCNS), and several techniques are available. In this study, we evaluated the outcomes of 2 different UCNS techniques performed in our(More)
This paper presents Cluster-Aware Remote Disks (CARDs), a Single System I/O architecture for cluster computing. CARDs virtualize accesses to remote cluster disks over a System Area Network. Their operation is driven by cooperative caching policies that implement a joint management of the cluster caches. All the CARDs of a given disk employ a common policy,(More)
This paper presents the design and performance of remote disk drivers for clusters of Commodity-Off-The-Shelf PCs that fetch disk blocks over System Area Networks. The driver offers a flexible interface, being capable to logically act either as computeror network-attached storage. It allows for fine-grain remote cache control through exclusive caching. An(More)
The TCP connection endpoint migration allows arbitrary server-side connection endpoint assignments to server nodes in cluster-based servers. The mechanism is client-transparent and supports back-end level request dispatching. It has been implemented in the Linux kernel and can be used as part of a policy-based software architecture for request distribution.(More)
This paper presents the design of a Web server using multi-core processors and the Stream Control Transmission Protocol (SCTP) as a transport-level protocol for HTTP. The multi-threaded server design takes advantage of the underlying multi-core architecture by defining stream scheduling policies that attempt to improve the performance of the server threads.(More)
This paper presents some of the issues related to the use of multi-core architectures for real-time systems, in particular for real-time Java. Currently, the Real-Time Specification for Java (RTSJ) does not account for aspects of parallel computing, but the widespread use of multi-core architectures advocates for the need of a change in this regard. Some of(More)
This paper presents speculative TCP connection admission, a mechanism for improving sub-optimal request distribution decisions in cluster-based servers. Overloaded server nodes in the cluster speculatively accept incoming requests only to offload them to less-loaded nodes. Speculative admission uses connection endpoint migration, a technique that allows(More)