When i was querying to hive using mr execution engine, some containers use more than 100% cpu.
I was see from 50% cpu usage up to 1000% cpu usage. (using top command on each node manager host)
I don't understand why one container uses more than 100% cpu.
is Map or Reduce working multithreaded?
From the 'top' documentation:
%CPU -- CPU Usage : The percentage of your CPU that is being used by the process. By default,
top displays this as a percentage of a single CPU. On multi-core systems, you can have percentages that are greater than 100%. For example, if 3 cores are at 60% use,
top will show a CPU use of 180%.
Yes, MapReduce utilizes more than one core on your machine - it is parallelized at the node level as well as at the process level to take advantage of as many cores as possible. The processing of each row of data is independent of all other rows of data so that the data can be split up in as many ways as you have processing capabilities.