如果我跑:
sbin/start-dfs.sh
然后,尽管打印,它实际上并没有启动名称节点:
Starting namenodes on [0.0.0.0]
0.0.0.0: starting namenode, logging to /home/myuser/apps/hadoop-2.7.4/logs/hadoop-myuser-namenode-mybox.out
localhost: starting datanode, logging to /home/myuser/apps/hadoop-2.7.4/logs/hadoop-myuser-datanode-mybox.out
Starting secondary namenodes [0.0.0.0]
0.0.0.0: starting secondarynamenode, logging to /home/myuser/apps/hadoop-2.7.4/logs/hadoop-myuser-secondarynamenode-mybox.out
BUT 如果我改为bin/hadoop namenode
,那么namenode就会启动。为什么呢?
答案 0 :(得分:0)
我假设你的机器上有hadoop-2。*,
sbin / start-dfs.sh
您的输出似乎没问题, 检查jps或尝试登录namenode url,
和 bin / hadoop namenode:
使用gedit或任何编辑器打开$ HADOOP_HOME / bin / hadoop文件, 验证它对参数 namenode
的作用通常用于格式化namenode,使用命令 hadoop或hdfs namenode -format。
用于启动dfs和yarn
<强> start-dfs.sh 强>
<强> start-yarn.sh 强>
将使用命令