Datanode cannot be started in hadoop2.0

Problem description: It can be started successfully when the datanode process is started. However, when you use JPs to view the process, it is found that the process does not exist. The following is the error message in the datanode log file.

As shown in the screenshot:

 Main reasons: Cause of error: the data in datanode cannot find the corresponding handle in the namenode after being formatted because the data is placed in the zero-time directory of TMP.

Solution:

1. First, delete the log information of all data under the logs/directory.

2. Delete all files in the temp file in the DFS directory (for hadoop configuration process, refer to"Hadoop2.20.0 cluster installation tutorial ")

3. reformat the namenode node.

 

 

Datanode cannot be started in hadoop2.0