如果YARN容器超出其堆大小设置，map或reduce任务将失败，并出现类似于以下的错误:2015-02-0611:58:15,461WARNorg.apache.hadoop.yarn.server.nodemanager.containermanager.monitor.ContainersMonitorImpl:Container[pid=10305,containerID=container_1423215865404_0002_01_000007]isrunningbeyondphysicalmemorylimits.Currentusage:42.1GBof42GBphys