MapReduce is an important programming model for building data centers containing ten of thousands of nodes. In a practical data center of that scale, it is a common case that I/Obound jobs and CPU-bound jobs, which demand different resources, run simultaneously in the same cluster. In the MapReduce framework, parallelization of these two kinds of job has not been concerned. In this paper, we give a new view of the MapReduce model, and classify the MapReduce workloads into three categories based on their CPU and I/O utilization. With workload classification, we design a new dynamic MapReduce workload predict mechanism, MR-Predict, which detects the workload type on the fly. We propose a Triple-Queue Scheduler based on the MR-Predict mechanism. The Triple-Queue scheduler could improve the usage of both CPU and disk I/O resources under heterogeneous workloads. And it could improve the Hadoop throughput by about 30% under heterogeneous workloads.

No abstract

No abstract

The statistical law that governs the drift velocity of tropical cyclones in the Northwest Pacific Ocean is investigated. The investigation is based on data published by China Meteorological Administration for historical tracks of 2146 cyclone events that occurred during 1949-2012. Empirical formulae are obtained to relate the magnitude, the direction, the meridional and zonal components of the averaged cyclone drift velocity with latitude. As the latitude effect is excluded, it is found that the cyclone drift velocity is governed by simple statistical laws, i.e. the magnitude and direction of the deviated drift velocity approximately satisfy a gamma distribution and a symmetric bimodal distribution, respectively, while the meridional and zonal components of the deviated drift velocity satisfy the same type of symmetric probability distribution represented by the hyperbolic secant function but with different deviations. The results obtained are potentially applicable to the enhancement of current tropical cyclone track forecasting techniques. They are also useful in risk management over the coastal areas where tropical cyclones may cause serious damages.

No abstract