3 was installed in vmware virtual machine. A, B, C, with A hadoop cluster 3.2.1, but using the start - DFS. Sh start cluster, the command line always stuck in "Starting the name nodes [C] A" no response, the JPS looked at, the namenode is up, but because here gets stuck, lead to the back of the datanode, ZKFC, journalnode etc all have, add A echo script, printed executive command is this: HDFS - workers - config/data/hadoop - ha/etc/hadoop - hostnames daemon start "A C" -- the namenode, I remove the - workers, simple implementation HDFS daemon start - the namenode is no problem, with SSH and PDSH all not line, have A great god gives directions what reason be?

CodePudding user response:

If there is any error message?

CodePudding user response:

Without any error, I found that I input: under the command line SSH & lt; Hostname> "Daemon start the namenode HDFS -" is the same (& lt; Hostname> Is my the name of the machine), the command line has not been returned as a result, but I used: SSH & lt; Hostname> "Cat text. TXT" this is no problem, immediately return results and exit to the command line, and I use SSH & lt; Hostname> "Daemon stop the namenode HDFS -" is no problem, is the time to start, has not returned,

CodePudding user response:

Don't blindly to find the reason, take a look at your log
Refer to the article, https://blog.csdn.net/dzh284616172/article/details/105980074