Hadoop Datanode failed to start error

In the process of starting Hadoop distributed deployment, we found that Datanode did not start properly, and looked at the log to find the error:

Java.io.IOException:Incompatible clusterids In/opt/hadoop-2.5/dfs/data:namenode Clusterid = cid-adf01a94-ae34-4313-acf9-3781a425de66; Datanode Clusterid = cid-e00fcbab-47c2-4e73-8a4b-c8754dc9960e

The reason is because Datanode's Clusterid and Namenode's Clusterid mismatch: After the first format of DFS, Hadoop was started and used, and then the Format command (HDFs Namenode-format) was re-executed. At this point the clusterid of Namenode is regenerated, and Datanode's clusterid remains unchanged.

Solution: Enter the/opt/hadoop-2.5/dfs/name/current/directory, edit the version text file,

After you restart DFS, change the value of Clusterid cid-adf01a94-ae34-4313-acf9-3781a425de66 to cid-e00fcbab-47c2-4e73-8a4b-c8754dc9960e The Datanode node returns to normal.

Hadoop Datanode failed to start error