Learn More
Most supercomputers nowadays are based on large clusters, which call for sophisticated, scalable, and decentralized metadata processing techniques. From the perspective of maximizing metadata throughput, an ideal metadata distribution policy should automatically balance the namespace locality and even distribution without manual intervention. None of(More)
Path visual coverage is the region which could be seen from the path in the terrain. This paper advances the concept of "average horizon " which represents the wide extent of the visual field of each path point to analyze and model the max and min path visual coverage problems, and resolves them utilizing simulated annealing algorithm based on the operation(More)
Distributed metadata servers are required for cluster file system's scalability. However, how to distribute the file system metadata among multiple metadata servers and how to make the file system reliable in case of server failures are two difficult problem. We present a journal-based failure-recovery mechanism for distributed metadata servers in the(More)
There have been several algorithms for mining association rules, such as Apriori and some improved Aprioris, only to be Interested in those itemsets, which have support above a userdefined threshold. However, there exists a kind of important rule, indirect association, hidden in these itemsets, which are filtered out. When a pair of items, (A, B), which(More)
The major challenge in designing cluster file systems is to provide high aggregate I/O bandwidth and high metadata processing throughput for applications running on large-scale cluster systems. And with the rapid increase of required data storage, how to improve access performance of a PB-scale cluster file system is also a challenging issue. In this paper,(More)
Automated fault diagnosis becomes increasingly important to satellite work. In an early paper, we introduced system level diagnosis theory into satellite iietwork firstly and presented two-level-node graph, a novel modeling method,. Based on these work, a new test invalidation model under certain fault pattern is presented and a diagnosis algorithni is(More)
It is more and more important in data mining field to finding the frequent sequences in a large database. The paper briefly introduces the basic concept of frequent sequence mining and presents the data parallel formulation and task parallel formulation of tree-projection based algorithm. Moreover, the on-line LPT algorithm is used to successfully solve the(More)
  • 1