启动集群报错

问题

启动HDFS时有报错, hadoop102上没有NameNode进程

[k@hadoop102 hadoop-3.3.6]$ sbin/start-dfs.sh 
Starting namenodes on [hadoop102]
Starting datanodes
localhost: mv: 无法获取'/opt/module/hadoop-3.3.6/logs/hadoop-k-datanode-hadoop102.out' 的文件状态(stat): 没有那个文件或目录
Starting secondary namenodes [hadoop104]
[k@hadoop102 hadoop-3.3.6]$ jps
15929 Jps
11711 DataNode
1
2
3
4
5
6
7
8

解决

[k@hadoop102 hadoop-3.3.6]$ vim etc/hadoop/workers 
localhost #删除改行可解决报错提示
hadoop102
hadoop103
hadoop104
1
2
3
4
5

重新格式化NameNode

1. 停掉所有进程
2. 删除每台机器hadoop-3.1.3下的data和logs
3. 在NameNode配置机器上格式化
4. 重新启动集群

[k@hadoop102 hadoop-3.3.6]$ rm -rf data/ logs/
[k@hadoop102 hadoop-3.3.6]$ hdfs namenode -format
[k@hadoop102 hadoop-3.3.6]$ sbin/start-dfs.sh 
Starting namenodes on [hadoop102]
Starting datanodes
hadoop103: WARNING: /opt/module/hadoop-3.3.6/logs does not exist. Creating.
hadoop104: WARNING: /opt/module/hadoop-3.3.6/logs does not exist. Creating.
Starting secondary namenodes [hadoop104]
[k@hadoop102 hadoop-3.3.6]$ jps
2914 NameNode  #有了√
3094 DataNode
3388 Jps
1
2
3
4
5
6
7
8
9
10
11
12