How Hadoop manage load balancing -
i have started working on hadoop.
want know how hadoop manage load balancing.
if have 5nodes in 1 cluster how hadoop ensure each node have equal work load?
there algorithm used hadoop load balancing?
could please me learn hadoop?
i'll assume mean yarn, resourcemanager, not hdfs, filesystem.
yarn not ensure nor guarantee equal processing.
in terms of mapreduce, if data heavily skewed towards particular key pairs, 1 process of 1 node bottleneck job.
if instead meant hdfs, there's literally called hdfs rebalancer, that's ensure data spread within cluster jobs can become better distributed in terms of "data locality". however, still won't skewness of data.
Comments
Post a Comment