高级检索

    负载平衡的多级并行对等计算在新药研发网格中的实现

    Realization of Multi-Level Parallel Load-Balance P2P Computation in the DDG

    • 摘要: 介绍了一种在Internet环境下利用分布的计算资源实现计算 /数据密集的高通量新药筛选的应用网格———新药研发网格 (drugdiscoverygrid ,DDG) 对DDG中采用的多级并行对等计算模型 (P2P) ,自适应的动态负载平衡算法及实现 ,网格资源的容错技术以及系统安全做了较详细的描述 ,给出了一组实际环境下的实验数据 实验证明 ,DDG具有很好的负载平衡能力和良好的系统健壮性 ,新药筛选应用在大规模的互连网环境下可以获得良好的并行加速效果

       

      Abstract: Described in this paper is a computing grid, the drug discovery grid (DDG), which utilizes the distributed resources to implement the efficient computing-intensive/data-intensive drug-design applications in the Internet.Discussed in detail are the multilevel parallel P2P technologies, the self-adaptive dynamical load balancing schedule technologies, the grid resources’ fault tolerance technologies, and the system security technologies.The experimental results show that the DDG’s load balance is effective and robust and the drug-design application has good parallel acceleration in the large-scale Internet environment.

       

    /

    返回文章
    返回