Hadoop Learning 17--yarn Configuration Chapter-Basic Configuration Node

<Configuration> 　　 < Property> 　　　　 <name>Yarn.nodemanager.aux-services</name> 　　　　 <value>Mapreduce.shuffle</value> 　　 </ Property> 　　 < Property> 　　　　 <name>Yarn.nodemanager.aux-services.mapreduce.shuffle.class</name> 　　　　 <value>Org.apache.hadoop.mapred.ShuffleHandler</value> 　　 </ Property></Configuration>

To be able to run a mapreduce program, you need to have each NodeManager load shuffle at startup Server,shuffle Server is actually Jetty/netty server,reduce The task uses the server to remotely copy the intermediate results from the map task from each NodeManager. The two additional configurations above are used to specify shuffle serve. If the yarn cluster has multiple nodes, you also need to configure parameters such as Yarn.resourcemanager.address

Hadoop Learning 17--yarn Configuration Chapter-Basic Configuration node