A slave node in your cluster has 24GB of Ram and 12 physical processor cores on hyper threading-enabled processor. You set the value of mapped. child.java.opts to -Xmx1G, and the value of…

Question

Options

• A24
• B
• C6
• D2
• E12
• F20

• A
  4% (1)
• B
  9% (2)
• C
  4% (1)
• E
  83% (19)

Explanation

There is 24 GB available. Each node will use 1 GB (child.java.opts to -Xmx1G). We should use a maximum of 24 nodes (12 mappers and 12 reducers). * Use mapred.tasktracker.map.tasks.maximum and mapred.tasktracker.reduce.tasks.maximum to control the maximum number of map/reduce tasks spawned simultaneously on a TaskTracker. * Hadoop should be configured correctly to get the best performance. If on the Analysis phase the hadoop machines go into a non responsive state (e.g., you cannot connect to them) and it's determined that this is because of very high CPU usage and it doesn't get out of that state for a few hours, this could be a hadoop configuration problem. The following formula should be applied to each hadoop node (master and slave) so that the machines never go into the non responsive mode: (number of mapper + number of reducers) * memory setting < available memory number of mappers + number of reducers <= available CPU's memory per mapper/reducer can not be less than 2GB

No community discussion yet for this question.