Administrate, optimize a Hadoop cluster


Module M15

The “Administration Optimization training of a Hadoop cluster” aims to train engineers and IS operators to administer Hadoop clusters. It is aimed at computer training populations (IT administrators) with a solid knowledge of Linux.


Program administrate, optimize a Hadoop cluster

  • Hadoop ecosystem quick reminders
  • Add users and give them access to cluster services and HDFS
  • Deal with the need for impersonalization
  • Monitor your cluster – the logs
  • Stop and restart Hadoop services
  • Tour of Hadoop properties important for optimization
  • Configure containers yarn
  • Configure the capacity scheduler
  • De-commissioning and re-commissioning nodes in the Hadoop cluster
  • Load balancing data when adding node
  • Troubleshooting : classic problems

Prerequisites : Module M1, Module M2 & Solid Knowledge of Linux