Page 29 - MCPW - Multicloud Presales Workshop
P. 29

Extract Information Map reduce







                                              Distributed Computing


                         MapReduce

                         A computing task is parallelized by distributing data

                        onto multiple worker nodes

                         The dataset cannot be stored on a single physical node

                        Data is stored local to the compute process


















                                                                       29
   24   25   26   27   28   29   30   31   32   33   34