Yet Another Resource Negotiator (YARN) DATA-SCIENCE-OPTIMIZE Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   How does YARN handle the failure of a NodeManager?

   • AThe failed tasks are lost and the application fails
   • BThe ResourceManager restarts the entire cluster
   • CThe NodeManager is automatically replaced with a new one
   • DThe ApplicationMaster is notified and can reschedule tasks on other nodes
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   What is the first step in the YARN application submission workflow?

   • AThe NodeManager starts containers for tasks
   • BThe ResourceManager launches the ApplicationMaster
   • CThe ApplicationMaster requests resources from the ResourceManager
   • DThe client submits the application to the ResourceManager
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A NodeManager on a YARN cluster fails and goes offline. What is the immediate impact on the applications running on that node?

   • AThe applications on that node will be paused until the NodeManager comes back online.
   • BAll containers running on that NodeManager are killed, and the ApplicationMasters of the affected applications will request new containers from the ResourceManager.
   • CThe containers continue running because they are managed by the ResourceManager, not the NodeManager.
   • DThe ResourceManager will restart the NodeManager automatically and resume the containers.
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   What does the ApplicationMaster do when a task fails during execution?

   • AIt shuts down the application immediately
   • BIt requests a new container and re-executes the task
   • CIt restarts the entire cluster
   • DIt ignores the failure and continues
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A MapReduce job is running on a YARN cluster. A NodeManager running one of the map tasks crashes due to a hardware failure. The ApplicationMaster detects that the map task has failed. What is the ApplicationMaster's immediate responsibility in this situation?

   • ARe-run the failed task locally on the ApplicationMaster's own container to avoid requesting new resources.
   • BRequest a new container from the ResourceManager and reschedule the failed map task on another available NodeManager.
   • CRestart the failed NodeManager and wait for the container to be re-launched on the same node.
   • DNotify the ResourceManager to kill the entire application and resubmit it from the client.
   Reveal Solution Discussion Select an answer first