Performance Analysis for Fast Parallel Recomputing Algorithm under DTA

摘要

With the rapid increasing of spatial data resolution, the huge volume of datasets makes the geo-computation more time-consuming especially in operating some complex algorithms. Parallel computing is regarded as an efficient solution by utilizing more computing resource. The stable and credible services play an irreplaceable role in parallel computing, especially when an error occurs in the large-scale science computing. In this paper, a master/slave approach of implementing the fast parallel recomputing is proposed based on redundancy mechanism. Once some errors in application layer are detected, the original data block with computation errors is further partitioned into several sub-blocks which are recomputed by the surviving processes concurrently to improve the efficiency of failure recovery. The multi-thread strategy in main process is adopted to distribute data block, detect errors and start recomputing procedure concurrently. The experimental results show that the proposed method can achieve better performance efficiency with fewer additional overhead.