在yarn 2.6.0集群中,有没有一种方法可以在resourcemanager上获得所有连接节点的cpu利用率?此外,源代码是否可修改,以便我们可以根据利用率来决定map reduce作业的节点。如果是,这种变化将在哪里发生?
68de4m5k1#
请找出容器监视器的实现:(cpu利用率)hadoop-2.6.0-src/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/monitor/containersmonitorimpl.java我们有办法检查集装箱是否超限。isprocesstreeoverlimit进程溢出 将向您展示如何获取特定容器(进程)的内存使用情况。hadoop-2.6.0-src/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/util/procfsbasedprocesstree.java上面的文件显示了yarn如何获取内存使用情况:跟踪/proc中的进程文件。
1条答案
按热度按时间68de4m5k1#
请找出容器监视器的实现:(cpu利用率)
hadoop-2.6.0-src/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/monitor/containersmonitorimpl.java
我们有办法检查集装箱是否超限。
isprocesstreeoverlimit进程溢出 将向您展示如何获取特定容器(进程)的内存使用情况。
hadoop-2.6.0-src/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/util/procfsbasedprocesstree.java
上面的文件显示了yarn如何获取内存使用情况:跟踪/proc中的进程文件。