Container sizes are determined by mapreduce.map.memory.mb and
mapreduce.reduce.memory.mb; both default to 1,024 MB. These settings are used by the
application master when negotiating for resources in the cluster, and also by the node manager, which runs and monitors the task containers. The heap size of the Java process is set by mapred.child.java.opts, and defaults to 200 MB. You can also set the Java options separately for map and reduce tasks (see Table 10-4).
AND
If a container uses more memory than it has been allocated, then it may be
terminated by the node manager and marked as failed.
0
0
Sudhasays:
I believe you have completed the admin certification.
Can you please send me the dump fro CCA 500 with the correct answer?
0
0
Bsays:
Do you have the dump? I’d love a copy of the correct answers as well.
0
0
testsays:
Do you have the dump with valid answers? Could you please share it
0
0
Trung Phansays:
Thanks Dev for your clarification.
0
0
ravsays:
Can you please update what is the answer?
0
0
Pathik Paulsays:
A – the question is about “per-application” Not Per host or not per data node.
Quote: “The ApplicationMaster is, in effect, an instance of a framework-specific library and is responsible for negotiating resources from the ResourceManager and working with the NodeManager(s) to execute and (!)monitor(!) the containers and their resource consumption.”
According to the fact that the question states “per-application resource using”, and from the documentation located at hortonworks, the answer should be A
0
0
abcsays:
Answer B.
client – submits the MapReduce job.
YARN resource manager – coordinates the allocation of compute resources on the cluster.
YARN node managers – launch and monitor the compute containers on machines in the cluster.
MapReduce application master – coordinates the tasks running the MapReduce job.
The application master and the MapReduce tasks run in containers that are scheduled by the resource manager and managed by the node managers.
0
0
hadoopexpresssays:
Answer is B.
The application master is not a daemon. It is created as a temporary task and vanishes when job completes.
Node manager is a daemon (as also Resource Manager, NameNode, DataNode). These run continuously and must be available as long as the cluster keeps running regardless of jobs submitted.
0
0
orangegallerysays:
It says either Daemon or service. Node Manager is limited to only Node level but application level. ApplicationMaster looks after the entire application across node managers. Answer should be A. Hope it makes sense.
0
0
Natarajan Ramanathansays:
Can someoone please provide the whole dump of the recent exams please.
0
0
Venkysays:
B. Node manager
Node Manager The Node Manager is responsible for managing each and every node within the YARN cluster. The Node Manager provides the services per-node within the cluster. These are variety of services ranging from monitoring the management of a container and its life cycle to monitoring the resources and keeping a track of the health and usage of resources of each node.
The answer is B, since per application resource usage is monitered by nodemaneger.
0
0
Its “A” , refer http://www.devx.com/opensource/intro-to-apache-mapreduce-2-yarn.html – Section Resource Manager,
0
0
Node Manager: from Definitive Guide:
Container sizes are determined by mapreduce.map.memory.mb and
mapreduce.reduce.memory.mb; both default to 1,024 MB. These settings are used by the
application master when negotiating for resources in the cluster, and also by the node manager, which runs and monitors the task containers. The heap size of the Java process is set by mapred.child.java.opts, and defaults to 200 MB. You can also set the Java options separately for map and reduce tasks (see Table 10-4).
AND
If a container uses more memory than it has been allocated, then it may be
terminated by the node manager and marked as failed.
0
0
I believe you have completed the admin certification.
Can you please send me the dump fro CCA 500 with the correct answer?
0
0
Do you have the dump? I’d love a copy of the correct answers as well.
0
0
Do you have the dump with valid answers? Could you please share it
0
0
Thanks Dev for your clarification.
0
0
Can you please update what is the answer?
0
0
A – the question is about “per-application” Not Per host or not per data node.
0
0
B –
https://www.mapr.com/blog/best-practices-yarn-resource-management
As per this site NodeManager can monitor the memory usage … application containers are killed by NodeManager. (sorry about the confusion)
0
0
Answer : B . ApplicationMaster only require resource from ResourceManager and does not monitor using of resource.
0
0
B
0
0
http://hortonworks.com/blog/apache-hadoop-yarn-concepts-and-applications/
Quote: “The ApplicationMaster is, in effect, an instance of a framework-specific library and is responsible for negotiating resources from the ResourceManager and working with the NodeManager(s) to execute and (!)monitor(!) the containers and their resource consumption.”
According to the fact that the question states “per-application resource using”, and from the documentation located at hortonworks, the answer should be A
0
0
Answer B.
client – submits the MapReduce job.
YARN resource manager – coordinates the allocation of compute resources on the cluster.
YARN node managers – launch and monitor the compute containers on machines in the cluster.
MapReduce application master – coordinates the tasks running the MapReduce job.
The application master and the MapReduce tasks run in containers that are scheduled by the resource manager and managed by the node managers.
0
0
Answer is B.
The application master is not a daemon. It is created as a temporary task and vanishes when job completes.
Node manager is a daemon (as also Resource Manager, NameNode, DataNode). These run continuously and must be available as long as the cluster keeps running regardless of jobs submitted.
0
0
It says either Daemon or service. Node Manager is limited to only Node level but application level. ApplicationMaster looks after the entire application across node managers. Answer should be A. Hope it makes sense.
0
0
Can someoone please provide the whole dump of the recent exams please.
0
0
B. Node manager
Node Manager The Node Manager is responsible for managing each and every node within the YARN cluster. The Node Manager provides the services per-node within the cluster. These are variety of services ranging from monitoring the management of a container and its life cycle to monitoring the resources and keeping a track of the health and usage of resources of each node.
0
0
It says per-application.
0
0
its says A-Application master
0
0
I choose A
0
0