esProc does not belong to Hadoop ecological system, and its parallel computing has nothing to do with MapReduce.
esProc aims at improving complex (semi)structured data computing, while Hadoop aims at mathematical algorithm of huge volume non-structured data.
esProc only focus on the computing jobs in (semi) structured data process, while hadoop has complete ecosystem and product line.
esProc can take Hadoop as data source to access HDFS and call Hive to control complex procedure, and assist Hadoop to accomplish complex procedure based computing.
esProc has more flexible and simpler parallel mechanism, and performance is close to or even better than database under small or mediate clusters. esProc can also achieve light-weight parallel commuting based on Hadoop.
No comments:
Post a Comment